As Of                Filer                Filing    For·On·As Docs:Size              Issuer               Agent

12/14/06  Cray Inc                          424B1                  1:1.0M                                   Bowne - Seattle/FA

Document/Exhibit                   Description                      Pages   Size 

 1: 424B1       Prospectus Pursuant to Rule 424(B)(1)               HTML    546K 

TABLE OF CONTENTS

RISK FACTORS

   An investment in our common stock involves a high degree of risk. You should carefully consider the risks and uncertainties described below together with all of the other information contained or incorporated by reference in this prospectus before deciding whether to invest. Each of these risks could adversely affect our business, financial condition and results of operations. As a result, the trading price of our common stock could decline and you might lose all or part of your investment.

Risk Factors Pertaining to Our Business, Operations and Industry

Our operating results may fluctuate significantly and we may not achieve profitability in any given period.

   Our operating results are subject to significant fluctuations due to the factors listed below, which make estimating revenue and operating results for any specific period very difficult. We experienced net losses in each full year of our development-stage operations prior to 2002. For 2002 we had net income of $5.4 million and for 2003 we had net income of $63.2 million, including an income tax benefit of $42.2 million substantially all of which came from the reversal of a valuation allowance against deferred tax assets. For 2004 we had a net loss of $207.4 million, including an expense for in-process research and development of $43.4 million and an income tax expense of $59.1 million, of which $58.9 million related to the establishment of a valuation allowance against deferred tax assets. For 2005 we had a net loss of $64.3 million, and for the first nine months of 2006 we had a net loss of $20.8 million.

   Whether we will be able to increase our revenue and achieve and sustain profitability on a quarterly and annual basis depends on a number of factors, including:

   The timing of orders and shipments impact our quarterly and annual results and are affected by events outside our control, such as:

   Because of the numerous factors affecting our revenue and results of operations, we cannot assure our investors that we will have net income on a quarterly or annual basis in the future. We anticipate that our quarterly results will vary significantly. Delays in product development, receipt of orders or product acceptances could have a substantial adverse effect on our results in 2006 and in future years.

Failure to sell Cray XT3 and Cray XT4 systems in planned quantities and at expected gross margins could adversely affect 2006 and 2007 revenue and operating results.

   We expect that a significant portion of our product revenue in the fourth quarter of 2006 and in 2007 will come from a limited number of sales of the Cray XT3 and Cray XT4 systems to governmental purchasers in the United States and overseas. We shipped the first Cray XT4 system in late November 2006, and thus the planned completion of customer acceptance late in the fourth quarter is at significant risk. We also face significant margin pressure for our Cray XT4 system and other commodity processor-based products from competitors. If we do not sell these systems in planned quantities and at expected gross margins, our 2006 and 2007 revenue and operating results would be adversely affected.

The achievement of our business plan is highly dependent on increased product revenue and margins.

   In 2005, we had lower revenue and margins than anticipated for our principal products. Product revenue was adversely affected by delays in product shipments due to development delays, including system software development for large systems, and at times by the availability of key components from third-party vendors. Potential system stability issues typical of new large systems could affect the timing of system acceptances, which would adversely affect our revenue, results of operations and cash flows. In the past, product margins have been adversely impacted by competitive pressures, lower volumes than planned and higher than anticipated manufacturing variances, including scrap, rework and excess and obsolete inventory. We sometimes do not meet all of the contract requirements for customer acceptance of our systems, which have resulted in contract penalties. Most often these penalties adversely affect the gross margin on a sale through the provision of additional equipment and services to satisfy delivery delays and performance shortfalls, although there is the risk of contract defaults and product return. The risk of contract penalties is increased when we bid for new business prior to completion of product development.

   To improve our financial performance, we need to receive higher margin orders, particularly for the Cray XT3 and Cray XT4 systems; deliver shipments and gain customer acceptances of new products on time, particularly the Cray XT4 system in the fourth quarter of 2006; and limit negative manufacturing variances, contract penalties and other charges that adversely affect product margin.

Our reliance on third-party suppliers poses significant risks to our business and prospects.

   We subcontract the manufacture of substantially all of the hardware components for all of our products, including integrated circuits, printed circuit boards, connectors, cables, power supplies and memory parts, on a sole or limited source basis to third-party suppliers. We use contract manufacturers to assemble our components for all of our systems. We also rely on third parties to supply key capabilities, such as file systems and storage subsystems. We are subject to substantial risks because of our reliance on limited or sole source suppliers. For example:

   Our products must meet demanding specifications. For example, integrated circuits must perform reliably at high frequencies to meet acceptance criteria. From time to time during 2004, 2005 and 2006, we incurred significant delays in the receipt of key components for the Cray X1E, Red Storm, Cray XT3, Cray XT4 and the Cray XD1 systems, which delayed product shipments and acceptances. The delays in product shipments and acceptances adversely affected 2004 and 2005 revenue and margins, and, to the extent that we experience similar problems in the future, such delays may adversely affect 2006 and future revenue and gross margins. We have also received parts that later proved defective, particularly for the Cray XD1 and Cray XT3 systems, which adversely affected our product and service margins and customer confidence.

   We have used IBM as a key foundry supplier of our integrated circuits for many years. In 2004 IBM informed us that it would no longer act as our foundry supplier on a long-term basis, although it will continue production of components for our current products for a limited time. We have negotiated a termination of the relationship with IBM and completed a general contract with Texas Instruments Incorporated (“TI”) to act as our foundry for certain key integrated circuits for our BlackWidow project in 2006 and 2007.

   Our Cray XT3, Cray XT4 and Cray XD1 systems utilize Advanced Micro Devices, Inc. (“AMD”) Opteron^tm processors as will planned upgrade and successor products. If Intel, IBM or other microprocessor suppliers develop processors with greater capabilities, even for a short time, our Cray systems, including upgrades and successor products, may be at a competitive disadvantage to systems utilizing such other processors. Our Cray XMT system is based on custom processors manufactured for us by Taiwan Semiconductor Manufacturing Company. If any of our integrated circuit suppliers suffers delays or cancels the development of enhancements to its processors, our

Phase III of the DARPA HPCS program will affect our operations.

   Our proposal for Phase III of the Defense Advanced Research Projects Agency (“DARPA”) High Productivity Computer Systems (“HPCS”) program was accepted by DARPA on November 21, 2006, for the development of our Cascade project. This award calls for the delivery of prototype systems by late 2010, and provides for a contribution by DARPA to us of up to $250 million payable over approximately four years, assuming we meet ten milestones. DARPA’s future financial commitments are subject to subsequent Congressional action. We will contribute at least $125 million towards the project’s total development cost. This award likely will result in increased net research and development expenditures by us for the cost-sharing portion of the program and may adversely affect our cash flow, particularly in the later years of the program.

We face last-time buy decisions affecting all of our current products, which may adversely affect our revenue and operating results.

   We have placed a last-time buy order for parts used to manufacture our Cray X1/ X1E products; we expect to no longer produce the Cray XD1 after 2006 and must plan our inventory purchases accordingly, and we face a last-time buy deadline in early 2007 for a key component for our Cray XT4 and Cray XMT systems and our BlackWidow project. Such last-time buy orders and inventory purchases must be placed before we know all possible sales prospects. In determining last-time buy orders and inventory purchases, we may either estimate low, in which case we limit the number of possible sales of products and reduce potential revenue, perhaps substantially, or we may estimate too high, and incur inventory obsolescence charges. Either way, our operating results would be adversely affected.

Our inability to overcome the technical challenges of completing the development of our supercomputer systems would adversely affect our revenue and operating results in 2006 and beyond.

   Our success in 2006 and in the following years depends in part on completing development of hardware and software enhancements to the Cray XT3 systems and successfully shipping and recognizing revenue for the Cray XT4 system in the fourth quarter of 2006. In 2007, we also must successfully and timely complete several key projects on our product roadmap, including the products based on our BlackWidow project and Cray XMT system. These hardware and software development efforts are lengthy and technically challenging processes, and require a significant investment of capital, engineering and other resources. Our engineering and technical personnel resources are limited. Unanticipated performance and/or development issues may require more engineers, time or testing resources than are currently available. Engineering resources directed to solving current issues may adversely affect the timely development of successor or future products. Given the breadth of our engineering challenges and our limited resources, we periodically review the anticipated contributions and expense of our product programs to determine their long-term viability. We may not be successful in meeting our development schedules for technical reasons and/or because of insufficient hardware and software engineering resources. Delays in successfully completing the design and production of the hardware components, including several custom integrated circuits and network components, delays in detecting and correcting, if possible, design errors in such integrated circuits and components and/or delays in developing requisite system software and needed software features and integrating and stabilizing the full systems would make it difficult for us to develop and market these systems timely and successfully and could cause a lack of confidence in our capabilities among our key customers. To the extent we uncover stability issues, whether for software or hardware, we may incur delays in passing acceptance tests, which would adversely affect our revenue, results of operations and cash flows, and may adversely affect the reputation of such systems in the market. Future sales of our products may be adversely affected. We have suffered

If we are unable to compete successfully in the HPC market, our revenue will decline.

   The performance of our products may not be competitive with the computer systems offered by our competitors. Many of our competitors are established companies well known in the HPC market, including IBM, NEC, Hewlett-Packard, Silicon Graphics, Inc. (“Silicon Graphics”), Dell, Bull S.A. and Sun Microsystems. Most of these competitors have substantially greater research, engineering, manufacturing, marketing and financial resources than we do.

   We also compete with systems builders and resellers of systems that are constructed from commodity components using microprocessors manufactured by Intel, AMD, IBM and others. These competitors include the previously named companies as well as smaller firms that benefit from the low research and development costs needed to assemble systems from commercially available commodity products. These companies have capitalized on developments in parallel processing and increased computer performance in commodity-based networking and cluster systems. While these companies’ products are more limited in applicability and scalability, they have achieved growing market acceptance. They offer significant peak/price performance on larger problems lacking complexity. Such companies, because they can offer high peak performance per dollar, can put pricing pressure on us in certain competitive procurements. In addition, to the extent that Intel, IBM and other microprocessor suppliers develop processors with greater capabilities than the processors we use from AMD, our Cray XT4 systems, including upgrades and successor products, may be at a competitive disadvantage to systems utilizing such other processors.

   Internationally we compete primarily with IBM, Hewlett-Packard, Sun Microsystems, Silicon Graphics and NEC. While the first four companies offer large systems based on commodity processors, NEC also offers vector-based systems with a large suite of ported application programs. As in the United States, commodity HPC suppliers can offer systems with significantly better peak/price performance. Periodic announcements by our competitors of new HPC systems or plans for future systems and price adjustments may reduce customer demand for our products. Many of our potential customers already own or lease very high performance computer systems. Some of our competitors may offer trade-in allowances or substantial discounts to potential customers, and engage in other aggressive pricing tactics, and we have not always been able to match these sales incentives. We have in the past and may again be required to provide substantial discounts to make strategic sales, which may reduce or eliminate any positive margin on such transactions, or to provide lease financing for our products, which could result in a deferral of our receipt of cash and revenue for these systems. These developments limit our revenue and resources and reduce our ability to be profitable.

   Our market is characterized by rapidly changing technology, accelerated product obsolescence and continuously evolving industry standards. Our success depends upon our ability to sell our current products, and to develop successor systems and enhancements in a timely manner to meet evolving customer requirements, which may be influenced by competitive offerings. We may not succeed in these efforts. Even if we succeed, products or technologies developed by others may render our products or technologies noncompetitive or obsolete. A breakthrough in technology could make low bandwidth cluster systems even more attractive to our existing and potential customers. Such a breakthrough would impair our ability to sell our products and reduce our revenue and operating results.

To be successful, we need to increase differentiation of our Cray XT4 and successor systems.

   We are a comparatively small company. We have concentrated our product roadmap on building balanced systems combining highly capable processors with very high speed interconnect and communications capabilities throughout the entire computing system. We achieve performance differentiation from our competitors through our custom processors in our vector-based and multithreading products, although the markets for those products may be limited in size. We need to establish greater performance differentiation from our competitors in our Cray XT3 and successor massively parallel products in order to command higher margins. The market for such products is larger but is replete with low bandwidth cluster systems offered by larger competitors with significant resources and smaller companies with minimal research and development expenditures. Potential customers may be able to meet their computing needs through the use of such systems, and are willing to accept lower capability and less accurate modeling in return for lower acquisition costs. Vendors of such systems, because they can offer high peak performance per dollar, put pricing pressure on us in certain competitive procurements. Our long-term success may be adversely affected if we are not successful in establishing the value of our balanced high bandwidth systems with the capability of solving challenging problems quickly to a market beyond our core of customers, largely certain agencies of the U.S. and other governments, that require systems with the performance and features we offer.

If the U.S. government purchases fewer supercomputers, our revenue would be reduced and our operating results would be adversely affected.

   Historically, sales to the U.S. government and customers primarily serving the U.S. government have represented a significant market for supercomputers, including our products. From January 1, 2001, through December 31, 2002, approximately 79%, of our product revenue was derived from sales to various agencies of the U.S. government; in 2003 and 2004, approximately 83% and 81%, respectively, of our product revenue was derived from such sales. In 2005, approximately 55% of our product revenue was derived from U.S. government sales, and in the first nine months of 2006, approximately 72% of our product revenue was derived from U.S. government sales. Our 2006 and future plans contemplate significant sales to U.S. government agencies. Sales to government agencies may be affected by factors outside our control, such as changes in procurement policies, budgetary considerations, domestic crises, and international political developments. If agencies and departments of the United States or other governments were to stop, reduce or delay their use and purchases of supercomputers, our revenue and operating results would be adversely affected.

If we lose government support for development of our supercomputer systems, our net research and development expenditures and capital requirements would increase and our ability to conduct research and development would decrease.

   A few government agencies and research laboratories fund a significant portion of our development efforts, including our BlackWidow, Cray XMT and Cascade projects, which significantly reduces our reported level of net research and development expenses. To date, our development contracts for our BlackWidow and Eldorado projects are not funded fully. Agencies of the U.S. government historically have facilitated the development of, and have constituted a market for, new and enhanced very high performance computer systems. U.S. government agencies may delay or decrease funding of our future product development efforts due to a change of priorities, international political developments, overall budgetary considerations or for any other reason. Any delay or decrease in other governmental support would cause an increased need for capital, increase significantly our research and development expenditures and adversely impact our operating results and our ability to implement our product roadmap.

We may infringe or be subject to claims that we infringe the intellectual property rights of others.

   Third parties in the past have asserted, and may in the future assert intellectual property infringement claims against us, and such future claims, if proved, could require us to pay substantial damages or to redesign our existing products or pay fees to obtain a cross-license agreement. Regardless of the merits, any claim of infringement would require management attention and could be expensive to defend.

Our indebtedness may adversely affect our financial strength.

   In December 2004 we sold $80.0 million in aggregate principal amount of our 3.0% Convertible Senior Subordinated Notes due 2024 (the “Notes”). Holders may require us to purchase all or a part of their Notes for cash at a purchase price of 100% of the principal amount of the Notes plus accrued and unpaid interest on December 1, 2009, 2014, and 2019, or upon the occurrence of certain events provided in the indenture governing the Notes. As of September 30, 2006, we had no other outstanding indebtedness for money borrowed and no material equipment lease obligations. We have a $30.0 million secured credit facility which supports the issuance of letters of credit and forward currency contracts. As of September 30, 2006, we had approximately $28.5 million available to borrow under this credit facility. The senior secured credit facility constitutes senior indebtedness with respect to the Notes. We may incur additional indebtedness for money borrowed, which may include borrowing under new credit facilities or the issuance of new debt securities. The level of our indebtedness could, among other things:

We will require a significant amount of cash to repay our indebtedness and to fund planned capital expenditures, research and development efforts and other corporate expenses.

   Our ability to make payments on our indebtedness, including the Notes, and to fund planned capital expenditures, research and development efforts and other corporate expenses will depend on our future operating performance and on economic, financial, competitive, legislative, regulatory and other factors. Many of these factors are beyond our control. Our business may not generate sufficient cash from operations and future borrowings may not be available to us in an amount sufficient to enable us to pay our indebtedness, including the Notes, or to fund our other needs.

   If we are unable to generate sufficient cash to enable us to pay our indebtedness, we may need to pursue one or more alternatives, such as reducing our operating expenses; reducing or delaying capital expenditures or research and development; selling assets; raising additional equity capital and/or debt; and seeking legal protection from our creditors.

   Any reduction in operating expenses, reduction or delay in capital expenditures, or sale of assets may materially and adversely affect our future revenue prospects. In addition, we may not be able to raise additional equity capital or debt on commercially reasonable terms or at all. Any of the above actions may not provide sufficient cash to repay our indebtedness, including the Notes. In addition, our issuance of additional equity or debt that is convertible into equity could dilute our existing shareholders.

If we cannot retain, attract and motivate key personnel, we may be unable to effectively implement our business plan.

   Our success also depends in large part upon our ability to retain, attract and motivate highly skilled management, technical, marketing, sales and service personnel. The loss of and failure to replace key engineering management and personnel could adversely affect multiple development efforts. Recruitment and retention of senior management and skilled technical, sales and other personnel is very competitive, and we may not be successful in either attracting or retaining such personnel. As part of our strategy to attract and retain personnel, we offer equity compensation through stock options and restricted stock grants. However, potential employees may not perceive our equity incentives as attractive, and current employees who have significant options with exercise prices significantly above current market values for our common stock may seek other employment. In addition, due to the intense competition for qualified employees, we may be required to increase the level of compensation paid to existing and new employees, which could materially increase our operating expenses.

Lower than anticipated sales of new supercomputers and the termination of maintenance contracts on older and/or decommissioned systems may reduce our service revenue and margins from maintenance service contracts.

   Our HPC systems are typically sold with maintenance service contracts. These contracts generally are for annual periods, although some are for multi-year periods, and provide a predictable revenue base. Our revenue from maintenance service contracts declined from approximately $95 million in 2000 to approximately $42 million in 2005. While we expect our maintenance service revenue to stabilize over the next year, we may have periodic revenue and margin declines as our older, higher margin service contracts are ended and newer, lower margin contracts are established, based on the timing of system withdrawals from service. Adding service personnel to new locations when we win contracts where we have previously had no presence and servicing installed products if we discover defective components in the field create additional pressure on service margins.

U.S. export controls could hinder our ability to make sales to foreign customers and our future prospects.

   The U.S. government regulates the export of HPC systems such as our products. Occasionally we have experienced delays for up to several months in receiving appropriate approvals necessary for certain sales, which have delayed the shipment of our products. Delay or denial in the granting of any required licenses could make it more difficult to make sales to foreign customers, eliminating an important source of potential revenue.

We may not meet the covenants imposed by our current credit agreement.

   We are subject to various financial and other covenants related to our line of credit with Wells Fargo Foothill, Inc. (“WFF”). If we were to fail to satisfy any of the covenants, we could be subject to fees and/or the possible termination of the credit facility. We failed to meet a financial covenant for 2005, which was waived by WFF. Termination of our credit facility could have an adverse impact on our liquidity.

The adoption of FAS 123R has and will continue to adversely affect our operating results and may adversely affect the market price of our common stock.

   We have used share-based compensation, primarily stock options and an employee stock purchase plan, as a key component in our employee compensation. We previously granted stock options to each new employee and to all employees on an annual basis. We believe we have structured these programs to align the incentives for employees with those of our long-term shareholders. We are reviewing our share-based compensation programs and their structure in light of the imposition of FAS 123R that became effective for us on January 1, 2006. In the past three fiscal years, as we have reported in the notes to our financial statements, our stock option program, as currently structured, would have added approximately $7 million to $26 million of additional non-cash expense annually. These estimates are based on use of the Black-Scholes valuation method. We recorded approximately $1.5 million as non-cash compensation expense in the first nine months of 2006 for stock options and unvested stock grants. We recently have granted some stock options to certain new employees with four-year vesting periods and have issued restricted stock grants to certain employees and officers, which will be recorded as an expense over the requisite service period. We do not know how analysts and investors will react to the additional expense recorded in our statement of operations rather than in the footnotes, and thus such additional expense may adversely affect the market price of our common stock.

We are required to evaluate our internal control over financial reporting under Section 404 of the Sarbanes-Oxley Act of 2002 at the end of each fiscal year, and any adverse results from such future evaluations could result in a loss of investor confidence in our financial reports and have an adverse effect on our stock price.

   Pursuant to Section 404 of the Sarbanes-Oxley Act of 2002, we are required to furnish a report by our management on our internal control over financial reporting in our Annual Report on Form 10-K. Such report must contain, among other items, an assessment of the effectiveness of our internal control over financial reporting as of the end of the fiscal year, including a statement as to whether or not our internal control over financial reporting is effective. This assessment must include disclosure of any material weaknesses in our internal control over financial reporting identified by management. Such report must also contain a statement that our independent registered public accounting firm has issued an attestation report on management’s assessment of such internal control. In our amended 2004 Annual Report on Form 10-K/ A, we identified and described a number of material weaknesses in our internal control over financial reporting, which required our assessment that our internal control over financial reporting was not effective, and our independent registered public accounting firm disclaimed an opinion with respect to our management’s assessment of our internal control over financial reporting as of December 31, 2004.

   Although we received favorable opinions from our independent registered public accounting firm and we reported no material weaknesses for 2005, we must continue to monitor and assess our internal control over financial reporting and determine whether we have any material weaknesses. Depending on their nature and severity, any future material weaknesses could result in our having to restate financial statements, could make it difficult or impossible for us to obtain an audit of our annual financial statements or could result in a qualification of any such audit. In such events, we could experience a number of adverse consequences, including our inability to comply with applicable reporting and listing requirements, a loss of market confidence in our publicly available information, delisting from the Nasdaq Global Market, loss of financing sources such as our line of credit, and litigation based on the events themselves or their consequences.

We incorporate software licensed from third parties into the operating systems for our products and any significant interruption in the availability of these third-party software products or defects in these products could reduce the demand for our products.

   The operating system software we develop for our HPC systems contains components that are licensed to us under “open source” software licenses. Our business could be disrupted if this software, or functional equivalents of this software, were either no longer available to us or no longer offered to us on commercially reasonable terms. In either case we would be required to redesign our operating system software to function with alternate third-party software, or develop these components ourselves, which would result in increased costs and could result in delays in product shipments. Furthermore, we might be forced to limit the features available in our current or future operating system software offerings. Our Cray XT4 and successor systems utilize software system variants that incorporate Linux technology. The SCO Group, Inc. has filed and threatened to file lawsuits against companies that operate Linux for commercial purposes, alleging that such use of Linux infringes The SCO Group’s rights. It is possible that The SCO Group could assert a claim of infringement against us with respect to our use of Linux technology. The open source licenses under which we have obtained certain components of our operating system software may not be enforceable. Any ruling by a court that these licenses are not enforceable, or that Linux-based operating systems, or significant portions of them, may not be copied, modified or distributed as provided in those licenses, would adversely affect our ability to sell our systems. In addition, as a result of concerns about The SCO Group’s lawsuit and open source generally, we may be forced to protect our customers from potential claims of infringement by The SCO Group or other parties. In any such event, our financial condition and results of operations may be adversely affected.

   We also incorporate proprietary software from third parties, such as for file systems, job scheduling and storage subsystems. We have experienced some functional issues in the past with implementing such software with our supercomputer systems. These issues, if repeated, may result in additional expense by us in integrating this software more fully and/or loss of customer confidence.

New environmental rules in Europe and other jurisdictions may adversely affect our operations.

   In 2006 members of the European Union (“EU”) and certain other European countries have begun implementing the Restrictions on Hazardous Substances (“RoHS”) Directive, which prohibits or limits the use in electrical and electronic equipment of the following substances: lead, mercury, cadmium, hexavalent chromium, polybrominated biphenyls, and polybrominated diphenyl ethers. After July 1, 2006, a company shipping products that do not comply with RoHS to the EU or such other European countries could have its products detained and could be subject to penalties. We decided not to ship any Cray X1E or Cray XD1 systems to Europe after July 1, 2006, because of these restrictions, and we are working with our suppliers to assure RoHS compliance with respect to our other products. We believe we are RoHS-compliant with our Cray XT4 system which began shipping in the fourth quarter of 2006. If a regulatory authority determines that one of our products is not RoHS-compliant, we will have to redesign and re-qualify certain components to meet RoHS requirements, which could result in increased engineering expenses, shipment delays, penalties and possible product detentions or seizures.

   A separate EU Directive on Waste Electrical and Electronic Equipment (“WEEE”) was scheduled to become effective in August 2005, but many EU member states have delayed its implementation. Under the WEEE Directive, companies that put electrical and electronic equipment on the EU market must register with individual member states, mark their products, submit annual reports, provide recyclers with information about product recycling, and either recycle their products or participate in or fund mandatory recycling schemes. In addition, some

   Other jurisdictions are considering adoption of rules similar to the RoHS and WEEE regulations. To the extent that any such rules differ from the RoHS and WEEE regulations, they may result in additional expense for us to redesign and qualify our products, and may delay us from shipping products into such jurisdictions.

In 2005 we formed a new senior management team that must work together effectively for us to be successful.

   In 2005 we revamped our senior management team, obtaining the services of Margaret A. Williams as Senior Vice President responsible for research and development, Brian C. Henry as Executive Vice President and Chief Financial Officer, Jan C. Silverman as Senior Vice President responsible for corporate strategy and business development and Steven L. Scott as Senior Vice President and Chief Technology Officer, and Peter J. Ungaro was elevated to Chief Executive Officer. We also added a new vice president/ corporate controller, a new vice president responsible for human resources, a new vice president responsible for hardware development and a new vice president responsible for software development. If our new management team, including any additional senior executives who join us in the future, is unable to work together effectively to implement our strategies, manage our operations and accomplish our business objectives, our ability to grow our business and successfully meet operational challenges could be significantly impaired. The loss of any key senior management could have a significant impact on our efforts to improve operating results.

We may not be able to protect our proprietary information and rights adequately.

   We rely on a combination of patent, copyright and trade secret protection, nondisclosure agreements and licensing arrangements to establish, protect and enforce our proprietary information and rights. We have a number of patents and have additional applications pending. There can be no assurance, however, that patents will be issued from the pending applications or that any issued patents will protect adequately those aspects of our technology to which such patents will relate. Despite our efforts to safeguard and maintain our proprietary rights, we cannot be certain that we will succeed in doing so or that our competitors will not independently develop or patent technologies that are substantially equivalent or superior to our technologies. The laws of some countries do not protect intellectual property rights to the same extent or in the same manner as do the laws of the United States. Additionally, under certain conditions, the U.S. government might obtain non-exclusive rights to certain of our intellectual property. Although we continue to implement protective measures and intend to defend our proprietary rights vigorously, these efforts may not be successful.

Risk Factors Pertaining to this Offering and Our Common Stock

Our stock price may be volatile or may decline regardless of our operating performance, and you may not be able to resell your shares at or above the public offering price.

   The stock market has been and is subject to price and volume fluctuations that particularly affect the market prices for small capitalization, high technology companies like us. The trading price of our common stock is subject to significant fluctuations in response to many factors, including our quarterly operating results (particularly if they are less than our or analysts’

A substantial number of our shares are eligible for future sale and may depress the market price of our common stock and may hinder our ability to obtain additional financing.

   As of September 30, 2006, we had outstanding:

   Almost all of our outstanding shares of common stock may be sold without substantial restrictions, with certain exceptions including 537,443 shares held by Board members, executive officers and key managers that may be forfeited and are restricted against transfer until vested. In addition, an aggregate of 538,012 shares beneficially owned by our executive officers and directors are subject to lock-up agreements with the underwriters for this offering and cannot be sold in the public market until the 91st day following the commencement of this offering, subject to a possible extension for up to 18 days in certain circumstances, as described below under “Underwriting.”

   Almost all of the shares of common stock that may be issued on exercise of the warrants and options will be available for sale in the public market when issued, subject in some cases to volume and other limitations. The warrants outstanding at September 30, 2006, consisted of warrants to purchase 50,000 shares of common stock, with an exercise price of $6.60 per share, expiring on June 3, 2009, and warrants to purchase 1,284,852 shares of common stock, with an exercise price of $10.12 per share, expiring on June 21, 2009. The Notes are not now convertible, and only become convertible upon the occurrence of certain events. We have registered the resale of the Notes and of the underlying common stock under the Securities Act of 1933, as amended (the “Securities Act”), which facilitates transferability of those securities. Sales in the public market of substantial amounts of our common stock, including sales of common stock issuable upon the exercise of warrants, options and Notes, may depress prevailing market prices for the common stock. Even the perception that sales could occur may impact market prices adversely. The existence of outstanding warrants, options and Notes may prove to be a hindrance to our future financings. Further, the holders of warrants, options and Notes may exercise or convert them for shares of common stock at a time when we would otherwise be able to obtain additional equity capital on terms more favorable to us. Such factors could impair our ability to meet our capital needs. We also have authorized 5,000,000 shares of undesignated preferred stock, although no shares of preferred stock currently are outstanding.

Our management will have broad discretion over the use of the net proceeds from this offering, and may invest or spend the net proceeds of this offering in ways with which you disagree.

   Our management will have broad discretion in the application of the net proceeds we receive from this offering and investors will be relying on the judgment of our management regarding the application of these proceeds. Management’s failure to apply these funds effectively could have an adverse effect on our ability to execute our business plan. In addition, the market price of our common stock may fall if the market does not view our use of the proceeds from the shares we are selling in this offering favorably.

Provisions of our Restated Articles of Incorporation and Bylaws could make a proposed acquisition that is not approved by our Board of Directors more difficult.

   Provisions of our Restated Articles of Incorporation and Bylaws could make it more difficult for a third party to acquire us. These provisions could limit the price that investors might be willing to pay in the future for our common stock. For example, our Restated Articles of Incorporation and Bylaws provide for:

   These provisions could delay, defer or prevent a merger, consolidation, takeover or other business transaction between us and a third party that is not approved by our Board of Directors.

SPECIAL NOTE REGARDING FORWARD-LOOKING STATEMENTS

   This prospectus, including information incorporated by reference, contains “forward-looking statements” within the meaning of Section 27A of the Securities Act and Section 21E of the Securities Exchange Act of 1934, as amended (the “Exchange Act”), and is subject to the safe harbor created by those sections and includes statements concerning the future of our industry, product development, business strategy, continued acceptance and growth of our products, dependence on significant customers and our expected future financial performance. These statements can be identified by the use of forward-looking terminology such as “may,” “will,” “expect,” “anticipate,” “estimate,” “continue,” or other similar words. When considering forward-looking statements, you should keep in mind the risk factors and other cautionary statements in this prospectus. The risk factors described above and other factors noted throughout this prospectus as well as factors that we do not now anticipate could cause our actual results to differ materially and adversely from those contained in any forward-looking statement.

   The risks, uncertainties and assumptions referred to above include the following: fluctuating operating results with possibility of periodic losses and uneven and possibly negative cash flows; the need for increased product revenue and margin, particularly from our Cray XT3, Cray XT4 and successor systems; the technical challenges of developing new supercomputer systems on time and budget; the timing of product orders, shipments and customer acceptances; the timing and level of government support for supercomputer system development; our dependency on third-party suppliers to build and deliver components timely that meet our specifications; the challenge of maintaining expense growth at modest levels while increasing revenue; our ability to attract, retain and motivate key employees, including executive officers and managers; and other risks that are described under ”Risk Factors” in this prospectus. Factors that could cause results to differ materially from those projected in the forward-looking statements are set forth in the discussion under “Risk Factors” above.

   In this prospectus, we rely on and refer to information and statistics regarding the markets for various products. We obtained this information from third party sources, discussions with our customers and our own internal estimates. We believe that these sources are reliable and estimates are reasonable, but we have not independently verified them. Because this information regardless of its source consists of estimates, it is likely that actual results will differ, perhaps materially, from such estimates.

USE OF PROCEEDS

   Our net proceeds from the sale of the 7,500,000 shares of our common stock offered by us at the public offering price of $10.00 per share are estimated to be $70.6 million after deducting the underwriting discounts and commissions and our estimated offering expenses.

   We have not paid cash dividends on our common stock and we do not anticipate paying any cash dividends on our common stock in the foreseeable future. In addition, our credit facility prohibits us from paying cash dividends without the consent of our lender.

   Our common stock is traded on the Nasdaq Global Market under the symbol “CRAY.”

   The last reported sale price of our common stock on the Nasdaq Global Market on December 13, 2006, was $10.45 per share. On December 11, 2006, we had 23,290,301 shares of common stock outstanding that were held by approximately 742 holders of record.

CAPITALIZATION

   The following table sets forth our actual cash and cash equivalents and capitalization as of September 30, 2006, and as adjusted to give effect to the issuance and sale by us of 7,500,000 shares of common stock in this offering at the public offering price of $10.00 per share, after deducting the underwriting discounts and commissions and estimated offering expenses payable by us.

   This capitalization table should be read in conjunction with our unaudited consolidated financial statements and related notes for September 30, 2006, which are incorporated by reference in this prospectus.

   The number of shares outstanding after the offering is based on 23,255,597 shares of common stock outstanding as of September 30, 2006, and excludes:

SELECTED CONSOLIDATED FINANCIAL DATA

   We derived the consolidated operating data for the years ended December 31, 2003, 2004 and 2005 and the consolidated balance sheet data as of December 31, 2004 and 2005 from our audited consolidated financial statements incorporated by reference in this prospectus. We derived the consolidated operating data for the fiscal years ended December 31, 2001 and 2002 and the consolidated balance sheet data as of December 31, 2001, 2002 and 2003 from our audited consolidated financial statements not incorporated by reference in this prospectus. We derived the consolidated balance sheet data as of September 30, 2006, and the consolidated operating data for the nine months ended September 30, 2005 and 2006 from our unaudited consolidated financial statements incorporated by reference in this prospectus. We have prepared the unaudited consolidated financial statements on a basis substantially consistent with the audited consolidated financial statements incorporated by reference in this prospectus and, in the opinion of management, these statements include all adjustments, consisting only of normal recurring adjustments, necessary for fair presentation of such data. The share and per share amounts reflect the one-for-four reverse stock split effected on June 8, 2006.

   The following data should be read in conjunction with “Management’s Discussion and Analysis of Financial Condition and Results of Operations” included elsewhere in this prospectus and the consolidated financial statements and related notes thereto incorporated by reference in this prospectus.

   In 2003 we recorded an income tax benefit of $42.2 million, principally as a result of the reversal of a $58.0 million valuation allowance against deferred tax assets.

   In 2004 the in-process research and development charge related to our acquisition of OctigaBay Systems Corporation. We recorded an income tax provision of $59.1 million, principally related to the establishment of a $58.9 million valuation allowance against deferred tax assets, primarily consisting of accumulated net operating losses.

QUARTERLY CONSOLIDATED FINANCIAL DATA

   The following table presents unaudited quarterly financial information for each of the eleven quarters ended September 30, 2006. In the opinion of management, this information contains all adjustments, consisting only of normal recurring adjustments, necessary for a fair presentation thereof. Certain 2004 and 2005 quarterly reclassifications have been made to conform to the 2006 presentation. The operating results are not necessarily indicative of results for any future periods. Quarter-to-quarter comparisons should not be relied upon as indicators of future performance. Our operating results are subject to quarterly fluctuations as a result of a number of factors. See “Risk Factors — Risk Factors Pertaining to Our Business, Operations and Industry.”

   The following data should be read in conjunction with “Management’s Discussion and Analysis of Financial Condition and Results of Operations” included elsewhere in this prospectus and the consolidated financial statements and related notes thereto incorporated by reference in this prospectus.

   The in-process research and development charge in the second quarter of 2004 related to our acquisition of OctigaBay Systems Corporation. The large net loss in the third quarter of 2004 included the establishment of a valuation allowance against our deferred tax asset of approximately $59 million.

   Since the second half of 2004, we have reviewed our workforce requirements in light of our operating results and have engaged in workforce reductions, particularly in the third quarter of 2004 and the second and fourth quarters of 2005. The 2005 fourth quarter also reflects a $4.9 million charge related to impairment of a core technology intangible asset.

   The decreases in research and development expenses beginning in the third quarter of 2005 reflect increased funding for our BlackWidow project and reduced expenses for the Cray XD1 system.

MANAGEMENT’S DISCUSSION AND ANALYSIS OF FINANCIAL CONDITION AND

   The following discussion should be read in conjunction with the consolidated financial statements and related notes thereto in our Annual Report on Form 10-K for the year ended December 31, 2005 (“2005 Form 10-K”) and the unaudited condensed consolidated financial statements and related notes thereto in our Quarterly Report on Form 10-Q for the fiscal quarter ended September 30, 2006 (“Third Quarter 2006 Form 10-Q”), both of which are incorporated by reference into this prospectus.

Overview

   We design, develop, manufacture, market and service high performance computing (“HPC”) systems, commonly known as supercomputers. Our supercomputer systems provide capability, capacity and sustained performance far beyond typical server-based computer systems and address challenging scientific and engineering computing problems.

   We believe we are well positioned to meet the HPC market’s demanding needs by providing superior supercomputer systems with performance and cost advantages when sustained performance on challenging applications and total cost of ownership are taken into account. We differentiate ourselves from our competitors primarily by concentrating our research and development efforts on the processing, interconnect and software capabilities that enable our systems to scale — that is, to continue to increase performance as our systems grow in size. Purpose-built for the supercomputer market, our systems balance highly capable processors, highly scalable software and very high speed interconnect and communications capabilities.

   In 2005, our management changed significantly with a new chief executive officer and new leaders in technology, engineering, finance, marketing, operations and customer support. Under our new management team, we have expanded our worldwide customer base, refined our product roadmap, established a lower operating cost model and sharpened our focus on execution to meet customer expectations and improve our financial operating results.

   Our management monitors and analyzes several key performance indicators in order to manage our business and evaluate our financial and operating performance, including:

   Revenue. Product revenue generally constitutes the major portion of our revenue in any reporting period, and for the reasons discussed elsewhere in this prospectus is subject to significant variability from period to period. In the short term, we closely review the status of product shipments, installations and acceptances in order to forecast revenue and cash receipts; longer-term, we monitor the status of the pipeline of product sales opportunities and pipeline and product development cycles. Revenue growth is the best indicator of whether we are achieving our objective of increased market share in the markets we address. Our new products scheduled for 2007 and our longer-term Adaptive Supercomputing vision are efforts to increase product revenue. Product revenue varies significantly from quarter to quarter. Service revenue is more constant in the short run and assists, in part, to offset the impact that the variability in product revenue has on total revenue.

   Gross margins. Our overall product margins in 2004 and 2005 were not satisfactory, even adjusting for the effect of our low-margin Red Storm and Cascade development projects, which were included as product revenue. To be successful, we need to increase product gross margins, which we believe is best achieved through increased product differentiation. We also monitor service margins and have been proactive in reducing service costs where possible. Our mid-term objective is to achieve overall margins, as a percentage of revenue, from 35% to 40% or better. Recent increases in gross margins have led to improved operational results.

   Operating expenses. Our operating expenses are driven largely by headcount, contracted research and development services and the level of co-funded research and development. We had two major headcount reductions in 2005. As part of our ongoing efforts to control operating expenses, we monitor headcount levels in specific geographic and operational areas. During 2006 we have been successful in receiving increased levels of co-funding for our research and development projects. Our recent Defense Advanced Research Projects Agency (“DARPA”) Phase III award is in line with our long-term development path. Our recent award likely will result in some increase in gross and net research and development expenditures by us in future periods due to the size of the overall program and the cost-sharing requirement on our part. Our overall operating expenses have significantly decreased in the first nine months of 2006 compared to the comparable 2005 period, especially in research and development. Our mid-term objective is to have total operating expenses, as a percentage of revenue, be in the range of 25% to 30%. Meeting this objective is dependent on our ability to grow revenue in the future.

   Liquidity and cash flows. Due to the variability in product revenue, our cash position also varies from quarter to quarter and within a quarter. We closely monitor our expected cash levels, particularly in light of potential increased inventory purchases for large system installations and the risk of delays in product shipments and acceptances and, longer term, in product development. This offering is consistent with our goal to build our cash position to provide additional working capital and to improve our operational and strategic flexibility while at the same time lowering the business risk to shareholders. Sustained profitability over annual periods is our primary objective, which should improve our cash position and shareholder value.

Critical Accounting Policies and Estimates

   This discussion, as well as disclosures included elsewhere in this prospectus, is based upon our consolidated financial statements, which have been prepared in accordance with generally accepted accounting principles (“GAAP”). The preparation of these financial statements requires us to make estimates and judgments that affect the reported amounts of assets, liabilities, revenue and expenses, and related disclosure of contingencies. In preparing our consolidated financial statements in accordance with GAAP, there are certain accounting policies that are particularly important. These include revenue recognition, inventory valuation, goodwill and other intangible assets, income taxes, the accounting for loss contracts and share-based compensation. Our relevant accounting policies are set forth in Note 3 to the consolidated financial statements of our 2005 Form 10-K and should be reviewed in conjunction with the condensed consolidated financial statements and notes as of September 30, 2006, in our Third Quarter 2006 Form 10-Q, as they are integral to understanding our results of operations and financial condition. In some cases, these policies represent required accounting. In other cases, they may represent a choice between acceptable accounting methods or may require substantial judgment or estimation.

   Additionally, we consider certain judgments and estimates to be significant, including those relating to the fair value allocation used in revenue recognition, percentage of completion accounting on the Red Storm contract, estimates of proportional performance on co-funded engineering contracts, determination of inventory at the lower of cost or market, useful lives for depreciation and amortization, determination of future cash flows associated with impairment testing for goodwill and long-lived assets, assumptions used to determine the fair value of stock options and assessments of fair value, estimation of restructuring costs, calculation of deferred income tax asset, potential income tax assessments and other contingencies. We base our estimates on historical experience, current conditions and on other assumptions that we believe to be reasonable under the circumstances. Actual results may differ from these estimates and assumptions.

   Our management has discussed the selection of significant accounting policies and the effect of judgments and estimates with the Audit Committee of our Board of Directors.

   We recognize revenue when it is realized or realizable and earned. In accordance with the Securities and Exchange Commission Staff Accounting Bulletin (“SAB”) No. 104, Revenue Recognition in Financial Statements, we consider revenue realized or realizable and earned when we have persuasive evidence of an arrangement, the product has been shipped or the services have been provided to our customer, the sales price is fixed or determinable, no significant unfulfilled obligations exist and collectibility is reasonably assured. In addition to the aforementioned general policy, the following are the specific revenue recognition policies for each major category of revenue and for multiple-element arrangements.

   Products. We recognize revenue from our product lines as follows:

   Revenue from contracts that require us to design, develop, manufacture or modify complex information technology systems to a customer’s specifications is recognized using the percentage of completion method for long-term development projects under American Institute of Certified Public Accountants (“AICPA”) Statement of Position 81-1, Accounting for Performance of Construction-Type and Certain Production-Type Contracts. Percentage of completion is measured based on the ratio of costs incurred to date compared to the total estimated costs. Total estimated costs are based on several factors, including estimated labor hours to complete certain tasks and the estimated cost of purchased components or services. Estimates may need to be adjusted from quarter to quarter, which would impact revenue and margins on a cumulative basis. To the extent the estimate of total costs to complete the contract indicates a loss, such amount is recognized in full in the period that the determination is made.

   Services. Maintenance services are provided under separate maintenance contracts with our customers. These contracts generally provide for maintenance services for one year, although some are for multi-year periods, often with prepayments for the term of the contract. We consider the maintenance period to commence upon installation and acceptance of the product, which may include a warranty period. We allocate a portion of the sales price to maintenance service revenue based on estimates of fair value. Revenue for the maintenance of computers is recognized ratably over the term of the maintenance contract. Maintenance contracts that are paid in advance are recorded as deferred revenue. We consider fiscal funding clauses as contingencies for the recognition of revenue until the funding is virtually assured. HPC service revenue is recognized as the services are rendered.

   Multiple-Element Arrangements. We commonly enter into transactions that include multiple-element arrangements, which may include any combination of hardware, maintenance and other services. In accordance with Emerging Issues Task Force Issue No. 00-21, Revenue Arrangements with Multiple Deliverables, when some elements are delivered prior to others in an arrangement

   If all of the criteria are not met, revenue is deferred until delivery of the last element as the elements would not be considered a separate unit of accounting and revenue would be recognized as described above under our product line or service revenue recognition policies. We consider the maintenance period to commence upon installation and acceptance of the product, which may include a warranty period and accordingly allocate a portion of the sales price as a separate deliverable which is recognized as service revenue over the entire service period.

   We record our inventory at the lower of cost or market. We regularly evaluate the technological usefulness and anticipated future demand of our inventory components. Due to rapid changes in technology and the increasing demands of our customers, we are continually developing new products. Additionally, during periods of product or inventory component upgrades or transitions, we may acquire significant quantities of inventory to support estimated current and future production and service requirements. As a result, it is possible that older inventory items we have purchased may become obsolete, be sold below cost or be deemed in excess of quantities required for production or service requirements. When we determine it is not likely we will recover the cost of inventory items through future sales, we write down the related inventory to our estimate of its market value. We are nearing the end of the life cycle for the Cray XT3 product and have made certain estimates of the future demand for this product. These estimates are subject to risk in the near term and could require a write-down of inventory if the actual demand is lower than currently estimated.

   Because the products we sell have high average sales prices and competitive product lives of generally one to two years, and because a high number of our prospective customers receive funding from U.S. or foreign governments, it is difficult to estimate future sales of our products and the timing of such sales. It also is difficult to determine whether the cost of our inventories will ultimately be recovered through future sales. While we believe our inventory is stated at the lower of cost or market and that our estimates and assumptions to determine any adjustments to the cost of our inventories are reasonable, our estimates may prove to be inaccurate. We have sold inventory previously reduced in part or in whole to zero, and we may have future sales of previously written down inventory. We also may have additional expense to write down inventory to its estimated market value. Adjustments to these estimates in the future may materially impact our operating results.

   Approximately 21% of our total assets as of September 30, 2006 consisted of goodwill resulting from our acquisition of the Cray Research business unit assets from Silicon Graphics in 2000 and our acquisition of OctigaBay Systems Corporation (“OctigaBay Systems”) in April 2004. We no longer amortize goodwill associated with the acquisitions, but we are required to conduct periodic analyses of the recorded amount of goodwill in comparison to its estimated fair value. We currently have one operating segment and reporting unit. As such, we evaluate any potential goodwill impairment by comparing our net assets against the market value of our outstanding shares of common stock. We performed an annual impairment test effective January 1, 2006, and determined that our recorded goodwill was not impaired.

   The analysis of whether the fair value of recorded goodwill is impaired and the number and nature of our reporting units involves a substantial amount of judgment. Future charges related to the amounts recorded for goodwill could be material depending on future developments and changes in technology and our business.

   In connection with our 2004 acquisition of OctigaBay Systems, we assigned $6.7 million of value to core technology. In December 2005 we announced plans to further integrate our technology platforms, and combine the Cray XD1 and the Cray XT3 products into a unified product offering. The expected undiscounted cash flows from the product using the core technology were not sufficient to recover the carrying value of the asset. We performed a fair value assessment similar to the original valuation and determined the asset had no continuing value. We wrote off the unamortized balance of our core technology intangible asset of $4.9 million. Accordingly, we recorded a $4.9 million charge in 2005 to “Restructuring, Severance and Impairment” in the Consolidated Statements of Operations appearing in our 2005 Form 10-K. In connection with this charge, we reversed the remaining deferred tax liability of $1.5 million that was established in the purchase accounting as amortization of this intangible asset was not deductible for income tax purposes.

   Deferred tax assets and liabilities are determined based on differences between financial reporting and tax bases of assets and liabilities and operating loss and tax credit carryforwards and are measured using the enacted tax rates and laws that will be in effect when the differences and carryforwards are expected to be recovered or settled. In accordance with Statement of Financial Accounting Standards (“FAS”) No. 109, Accounting for Income Taxes, a valuation allowance for deferred tax assets is provided when we estimate that it is more likely than not that all or a portion of the deferred tax assets may not be realized through future operations. This assessment is based upon consideration of available positive and negative evidence, which includes, among other things, our most recent results of operations and expected future profitability. We consider our actual historical results to have stronger weight than other more subjective indicators when considering whether to establish or reduce a valuation allowance on deferred tax assets.

   The provision for or benefit from income taxes represents taxes payable or receivable for the current period plus the net change in deferred tax assets and liabilities and valuation allowance amounts during the period. In 2003, we reversed $58.0 million of the valuation allowance against deferred tax assets, principally U.S. loss carryforwards, based primarily upon our consideration of our most recent profitable operating performance as well as our reasonably expected future performance. Based upon our judgment of the positive and negative evidence, we concluded that we would more likely than not be able to utilize most of our net deferred tax asset. In late 2004, we established a valuation allowance and recorded an income tax expense of $58.9 million based on our losses from operations in 2004 and based on our revised projections indicating continued challenging financial results. Based upon our most recent negative operating results, which we consider as a strong indicator of our future ability to utilize our deferred tax assets, we established a valuation allowance on certain deferred tax assets, principally U.S. loss carryforwards, created during 2005 in accordance with FAS No. 109.

   As of September 30, 2006, we had approximately $140.8 million of deferred tax assets, of which $140.1 million was fully reserved. The net deferred tax assets were generated in foreign jurisdictions where we believe it is more likely than not that we will realize these assets through future operations. For the nine month periods ended September 30, 2006, and 2005 we recognized income tax expense of $748,000 and $428,000, respectively. Income tax expense in all periods was related to taxes due in foreign and certain state jurisdictions.

   In accordance with our revenue recognition policy, certain production contracts are accounted for using the percentage of completion accounting method. We recognize revenue based on a measurement of completion comparing the ratio of costs incurred to date with total estimated costs multiplied by the contract value. Inherent in these estimates are uncertainties about the total cost to complete the project. If the estimate to complete results in a loss on the contract, we will record the amount of the estimated loss in the period the determination is made. On a regular basis, we update our estimates of total costs. Changes to the estimate may result in a charge or benefit to operations. As of September 30, 2006, our estimate of loss on the Red Storm contract was consistent with our estimate of such loss as of December 31, 2005, which was a cumulative loss of $15.3 million, all of which was recorded in prior periods. As of September 30, 2006 and December 31, 2005, the balance in the Red Storm loss contract accrual account was $3.0 million and $5.7 million, respectively, and is included in “Other accrued liabilities” in our consolidated balance sheets.

   On January 1, 2006, we adopted the fair value recognition provisions of Financial Accounting Standards Board (“FASB”) Statement No. 123(R), Share-Based Payment (“FAS 123R”). Prior to January 1, 2006, we accounted for share-based payments under the recognition and measurement provisions of APB Opinion No. 25, Accounting for Stock Issued to Employees (“APB 25”), and related Interpretations, as permitted by FASB Statement No. 123, Accounting for Stock-Based Compensation (“FAS 123”). In accordance with APB 25, no compensation cost was required to be recognized for options granted that had an exercise price equal to the market value of the underlying common stock on the date of grant. Certain of the stock options granted in connection with the OctigaBay Systems acquisition in 2004 had exercise prices below the fair market value of our common stock at the grant date and accordingly we have recorded compensation expense over the vesting period based on the intrinsic value method.

   We adopted FAS 123R using the modified-prospective transition method. Under that transition method, compensation cost recognized for the nine months ended September 30, 2006, includes: (a) compensation cost for all share-based payments granted prior to, but not yet vested, as of January 1, 2006, based on the grant-date fair value estimated in accordance with the original provisions of FAS 123, and (b) compensation cost for all share-based payments granted subsequent to January 1, 2006, based on the grant-date fair value estimated in accordance with the provisions of FAS 123R. The financial results for the prior periods have not been restated.

   Estimates of fair value of stock options are based upon the Black-Scholes option pricing model. We utilize assumptions related to stock price volatility, stock option term and forfeiture rates that are based upon both historical factors as well as management’s judgment.

   In June 2006, the FASB issued FASB Interpretation No. 48, Accounting for Uncertainty in Income Taxes (“FIN 48”). FIN 48 clarifies the accounting for uncertainty in income taxes and prescribes a recognition threshold and measurement attribute for the financial statement recognition and measurement of a tax position taken or expected to be taken in a tax return. FIN 48 is effective for financial statements as of January 1, 2007. We have not yet determined the impact of applying FIN 48.

   In September 2006, the FASB issued Statement of Financial Accounting Standards No. 157, Fair Value Measurements (“FAS 157”). FAS 157 defines fair value, establishes a framework for measuring fair value and expands disclosures about fair value measurements but does not require any new fair value measurements. FAS 157 is effective for financial statements issued for fiscal years beginning after November 15, 2007, and interim periods within those fiscal years. We have not yet determined the impact of applying FAS 157.

   In September 2006, the FASB issued Statement of Financial Accounting Standards No. 158, Employers’ Accounting for Defined Benefit Pension and Other Postretirement Plans (“FAS 158”). FAS 158 requires an employer to recognize the overfunded or underfunded status of a defined benefit postretirement plan (other than a multiemployer plan) as an asset or liability in its statement of financial position and to recognize changes in that funded status in the year in which the changes occur through comprehensive income. FAS 158 is effective for financial statements as of December 31, 2006. We have not yet determined the impact of applying FAS 158.

Results of Operations

   Our revenue, cost of revenue and gross margin for the years ended December 31, 2003, 2004 and 2005, and the nine months ended September 30, 2005, and 2006, were (in thousands, except for percentages):

  Product Revenue

   Product revenue for the nine months ended September 30, 2006, and 2005 consisted of $58.9 million and $84.5 million, respectively, primarily from Cray X1/X1E systems, Cray XT3 systems, Cray XD1 systems and other products, as well as $19.1 million and $15.3 million, respectively, from our Red Storm and Cascade development projects.

   The increase in 2005 product revenue over 2004 levels was due to increased sales of all three of our principal products, the Cray X1E, the Cray XT3 and the Cray XD1 systems, which became available after a production ramp-up during the first half of the year. In 2005, we recorded approximately $22.1 million in product revenue from the Cascade and Red Storm development projects which was a reduction of $27.4 million compared to 2004 due to reduced expenditures and associated revenue, in particular on the Red Storm development project. Product revenue declined in 2004 from 2003 due to significantly reduced Cray X1 system sales and the inability to recognize revenue on certain Cray X1E and Cray XT3 systems delivered in the fourth quarter of 2004, offset in part by an increase in Red Storm and Cascade development project revenue of $29.9 million to $49.5 million for 2004.

   We expect that product revenue for the fourth quarter of 2006 will improve significantly from the third quarter of 2006. There is a reasonable possibility that some or all of approximately $40 million of product revenue we expect to recognize in the fourth quarter of 2006 could be recognized in the first quarter of 2007 due to the timing of customer acceptances. If this were to occur, 2006 product revenue would be lower and 2007 product revenue would be positively impacted. We also will recognize approximately $39 million of revenue from one customer, Korea Meteorological Administration, in the fourth quarter of 2006.

   For the full year 2007, we expect strong product sales growth, offset in part by a nearly $20 million reduction in low margin development-related project revenue. The 2007 revenue level is dependent on the successful introduction of three new products over the next twelve months (our Cray XT4, our Cray XMT and code-named BlackWidow systems). Due in part to the timing of these product introductions, the first quarter of 2007 is expected to be the weakest quarter of 2007 and the second half of 2007 should be stronger than the first half of 2007.

  Service Revenue

   Service revenue for the nine months ended September 30, 2006 increased $5.6 million, or 16%, over the same period in 2005, due to a growth in maintenance revenue from new contracts.

   Service revenue in 2005 decreased slightly from 2004 due to lower revenue on maintenance contracts as older systems were withdrawn from service. Revenue from professional services in 2005 increased by $2.8 million from $3.7 million in 2004 due principally to a professional services contract to refurbish certain components for a customer. Service revenue decreased in 2004 from 2003 due to expiring maintenance contracts as older systems were withdrawn from service.

   While we expect our maintenance service revenue to stabilize and potentially increase over the next year, we may have periodic revenue and margin declines as our older, higher margin service contracts end. Our newer products will likely require less hardware maintenance and therefore generate less maintenance revenue than our historic vector systems. Overall service revenue may decline in 2007 due to the refurbishment professional services contract ending in 2006.

  Product Gross Margin

   Product gross margin improved 22 percentage points for the nine month period ended September 30, 2006, compared to the same period in 2005. This improvement in product gross margin was due to increased gross margins across all product lines, including lower charges for excess and obsolete inventory and no amortization of core technology intangible asset that was written off during the fourth quarter of 2005. Additionally, gross margins for the nine-month period ended September 30, 2005 were negatively impacted by a $7.7 million loss on the Red Storm project.

   Product gross margin in 2005, although improved compared to 2004, was impacted by several factors, including higher sales of the lower margin Cray XD1 product, a $7.7 million charge for a change in the estimate to complete the Red Storm project due principally to the addition of hardware deliverables to settle contract and performance issues, and $5.8 million of charges for inventory write-downs, which included scrap and obsolete inventory. Product gross margin in 2004 was negatively impacted by inventory write-downs of $8.5 million, a $7.6 million charge to record the initial estimated loss on the Red Storm project, a $1.6 million adjustment for unabsorbed manufacturing overhead relating to lower than planned production of Cray X1 systems and significant revenue recognized on the low margin Cascade and zero margin Red Storm projects. The Red Storm and Cascade research and development costs totaling $28.6 million, $57.3 million and $18.7 million in 2005, 2004 and 2003, respectively, are reflected on our financial statements as cost of product revenue and the related reimbursements are recorded in our financial statements as product revenue. Revenue for 2005, 2004 and 2003 included $2.1 million, $498,000 and $316,000, respectively, from the sale of obsolete inventory recorded at

   Product gross margin will be adversely affected in the fourth quarter of 2006 by the relatively low margin on a large contract expected to be recognized as revenue. With minimal low-margin development-related product revenue expected in 2007 and the expected benefit of three new product introductions, overall product gross margins should increase in 2007 as compared to 2006.

  Service Gross Margin

   Service gross margin improved 8 percentage points for the nine months ended September 30, 2006, compared to the respective 2005 period due to an increase in maintenance revenue while essentially maintaining costs at 2005 levels.

   In both 2005 and 2004, our service gross margin was favorably impacted by high margin professional service contracts, service cost reductions implemented in the fourth quarter of 2003 and the second half of both 2004 and 2005, and the completed amortization of legacy spare parts inventory by March 31, 2004, offset in 2005 in part by increased costs incurred to achieve customer acceptances of large Cray XT3 systems.

   We expect service gross margin for the fourth quarter of 2006 to be between 38% to 45%. Service gross margin percentage for 2007 is expected to be down somewhat as revenue from high margin professional services is expected to decrease.

   Our research and development expenses for the years ended December 31, 2003, 2004 and 2005, and the nine months ended September 30, 2005, and 2006, were (in thousands, except for percentages):

   Gross research and development expenses in the table above reflect all research and development expenditures, including expenses related to our research and development activities on the Red Storm and Cascade projects. Research and development expenses on our Red Storm and Cascade projects are reflected on our financial statements as cost of product revenue, and government co-funding on our other projects are recorded on our financial statements as reimbursed research and development. Research and development expenses include personnel expenses, depreciation, allocations for certain overhead expenses, software, prototype materials and outside contracted engineering expenses.

   Net research and development expenses in the first nine months of 2006 and in 2005 primarily reflect our net costs associated with the hardware and software associated with the Cray X1E, Cray XT3 and Cray XD1 systems, and the BlackWidow and Cray XMT projects, and their upgrade

   We have received increased government funding each period. For the nine months ended September 30, 2006, net research and development expenses decreased as compared to the same periods in 2005 due principally to increased funding for our BlackWidow project and reduced research and development expenses for the Cray XD1 product line.

   In 2005, net research and development expenses as a percentage of total revenue decreased as compared to 2004 due to higher revenue, increased government funding for our BlackWidow project, and the effect of a pay reduction program in the second half of 2005, partially offset by option expense as we accelerated vesting on options issued in connection with the OctigaBay Systems acquisition at the beginning of the second quarter in 2004. The higher 2004 percentage of total revenue compared to 2003 was due to lower revenue earned in 2004 and to increases in research and development expenses for most of our products and projects, including an increase of approximately $2.0 million to $2.5 million per quarter due to the OctigaBay Systems acquisition.

   With our recent DARPA Phase III award, net research and development expenses for the fourth quarter of 2006 will be lower than the third quarter of 2006. We anticipate 2007 net research and development expenses to be up slightly from 2006 due to the cost-sharing portion of the DARPA award. Net research and development expenses may also be higher in 2007, particularly in the second half, if the U.S. government ceases co-funding earlier than anticipated on our BlackWidow or Cray XMT projects. We expect that research and development co-funding, including amounts earned under the DARPA Phase III, BlackWidow and Cray XMT funding agreements, will be recorded primarily as a reduction to research and development expense in the fourth quarter of 2006 and in 2007.

   Our sales and marketing, general and administrative, and restructuring, severance and impairment charges for the years ended December 31, 2003, 2004 and 2005, and the nine months ended September 30, 2005, and 2006, were (in thousands, except for percentages):

   Sales and Marketing. The decrease in sales and marketing expenses for the nine month period ended September 30, 2006, compared to the same period in 2005, was primarily due to a decrease in headcount as a result of a reduction-in-force that took place in the second quarter of 2005.

   The decrease in 2005 sales and marketing expenses compared to 2004 was due to lower headcount, the pay reduction program in the second half of 2005 and lower discretionary spending, offset in part by higher commissions on increased product revenue. The increase in 2004 compared to 2003 was primarily due to an acceleration of expenses related to certain prepaid computer access services no longer utilized and additional benchmarking, application and sales

   We expect sales and marketing expenses to increase in the fourth quarter of 2006 due to increased sales commissions as a result of higher anticipated revenue and expenses related to a major supercomputing trade show, but decline for 2006 compared to annual 2005 levels due to reduced headcount, offset in part by re-establishment of full salaries for all of 2006. We expect that 2007 sales and marketing expenses will be slightly higher than 2006 levels primarily due to increased sales commissions on higher product sales.

   General and Administrative. The increase in general and administrative costs for the nine months ended September 30, 2006 over the corresponding 2005 period was primarily due to an increase in non-cash stock-based compensation incurred in connection with restricted stock and stock option grants as well as an increase in expense for variable pay and retention compensation expenses, which were partially offset by a general headcount decrease and the effects of the reduction-in-force that occurred in the second quarter of 2005.

   The decrease in general and administrative expense in 2005 compared to 2004 was primarily due to the effects of our reduction-in-force, as well as the pay reduction program in the second half of 2005, savings from which were offset in part by increased fees for external audit, Sarbanes-Oxley compliance and legal fees. The increase in 2004 compared to 2003 was due primarily to consulting costs related to Sarbanes-Oxley compliance and additional expenses as a result of our acquisition of OctigaBay Systems, including additional depreciation, insurance and utilities.

   We expect general and administrative expenses to remain modestly higher for the fourth quarter of 2006 compared to 2005 levels due to re-establishment of full salaries and our variable pay program as well as the executive retention and restricted stock grant programs instituted in December 2005. We expect 2007 general and administrative expenses to be similar to 2006 expense levels.

   Restructuring, Severance and Impairment. Restructuring, severance and impairment charges include costs related to our efforts to reduce our overall cost structure by reducing headcount. During the second quarter of 2005, we implemented a worldwide reduction in workforce of approximately 90 employees, or 10% of our worldwide workforce, primarily in manufacturing, sales, service and marketing. We incurred additional severance charges primarily for the retirement of our former Chief Executive Officer, James Rottsolk, in the third quarter of 2005. In the fourth quarter of 2005, we implemented an additional reduction of approximately 65 employees from our international sales, service and engineering operations, largely based in our Burnaby, British Columbia, Canada facility with the remainder in Europe. In the second quarter of 2006, we incurred additional severance costs related to both our second and fourth quarter 2005 actions; these costs decreased to $3,000 in the third quarter of 2006.

   In connection with the 2004 acquisition of OctigaBay Systems, we allocated $6.7 million of the purchase price to a core technology intangible asset, which was associated with the Cray XD1 system. In connection with the fourth quarter 2005 decision to incorporate the Cray XD1 system technology into the Cray XT3 line, as well as limited expected future benefits of the core technology obtained in the acquisition, we evaluated the carrying value of the unamortized balance of the intangible asset of $4.9 million and determined that the carrying value of the asset was impaired and accordingly recorded a charge for the $4.9 million in the fourth quarter of 2005.

In-Process Research and Development Charge

   As part of the acquisition of OctigaBay Systems, we incurred an expense associated with acquired in-process research and development of $43.4 million in the second quarter of 2004.

Other Income (Expense), Net

   For the nine months ended September 30, 2006 and 2005, we recognized net other expense of $2.0 million and $602,000, respectively. Other expense for the nine months ended September 30, 2006, was principally the result of a $1.6 million loss in fair value on our foreign currency forward contract derivative through September 30, 2006, while net other expense for the nine months ended September 30, 2005 was principally the result of net foreign currency transaction losses. We expect the impact of the foreign currency loss recorded in 2006 related to the foreign currency derivative will be recovered when the revenue on the related product sale is recognized.

   During 2005, we recorded $1.4 million of net other expense, compared to net other expense of $699,000 in 2004 and net other income of $1.5 million in 2003. Other expense in 2005 and 2004 primarily consisted of foreign currency losses on the remeasurement of foreign currency balances, principally intercompany balances. Other income in 2003 primarily consisted of gains resulting from the remeasurement of foreign currency balances, principally intercompany balances.

   Our interest income and interest expense for the years ended December 31, 2003, 2004 and 2005, and the nine months ended September 30, 2005, and 2006, were (in thousands):

   Interest income increased in the nine months ended September 30, 2006 compared to the same period in 2005 as a result of higher average invested cash balances and higher short-term interest rates.

   Interest income in 2005 and 2004 was related primarily to our cash and short-term investments balances, which, on average, were consistent with the balances during 2003. The 2003 interest income reflects our increased average cash position in 2003 following our public offering in February 2003 in which we raised $49.1 million.

   Interest expense for both nine month periods ended September 30, 2006, and 2005 principally consisted of $1.8 million of interest on our Notes in each period and $1.0 million and $672,000, respectively, of non-cash amortization of capitalized issuance costs. The amount of non-cash amortization of capitalized issuance costs was higher in 2006 than in 2005 because the line of credit agreement was entered into in May 2005.

   Interest expense in 2005 represents $2.4 million of interest on our Notes, $1.0 million of non-cash amortization of fees capitalized in connection with both our line of credit with WFF and our long-term debt offering costs, and $765,000 of interest and related fees on our line of credit with WFF. The interest expense for 2004 reflects approximately one month of interest on our Notes, one month of amortization of the related capitalized issuance costs and interest on our capital leases. The interest expense for 2003 reflects interest on our term loan for the first four months of the year and interest on our capital leases.

   We recorded tax expense of $748,000 for the nine months ended September 30, 2006, and $428,000 for the nine months ended September 30, 2005. The tax expense recognized in the first

   Benefit from income taxes in 2005 was $1.5 million, which consisted of a $2.3 million benefit for foreign deferred taxes, partially offset by current tax expense for local, state and foreign tax jurisdictions. We recorded an income tax provision of $59.1 million in 2004, principally related to the establishment of a $58.9 million valuation allowance against deferred tax assets, primarily consisting of accumulated net operating losses. Under the criteria set forth in FAS No. 109, Accounting for Income Taxes, management concluded that it was unlikely that the future benefits of these deferred tax assets would be realized. In 2003, we recorded an income tax benefit of $42.2 million, principally as a result of the reversal of $58.0 million valuation allowance on deferred tax assets.

   There has been no current provision for U.S. federal income taxes for any period. We have income taxes currently payable due to our operations in certain foreign countries, particularly in Canada and certain European and Asian countries and in certain states.

   As of December 31, 2005, we had tax net operating loss carryforwards of approximately $288 million that will begin to expire in 2010 if not utilized.

Liquidity and Capital Resources

   Cash, cash equivalents and accounts receivable totaled $90.0 million as of September 30, 2006, compared to $101.1 million as of December 31, 2005; cash and cash equivalents decreased by $1.9 million while accounts receivable decreased by $9.2 million. As of September 30, 2006, we had working capital of $42.7 million compared to $52.2 million as of December 31, 2005.

   Net cash used by operating activities for the nine months ended September 30, 2006 was $2.2 million compared to a use of $60.1 million for the same period in 2005. For the nine months ended September 30, 2006, cash used by operating activities was principally the result of our net loss for the period and an increase in inventory, partially offset by a decrease in accounts receivable and increases in deferred revenue and accounts payable. For the nine months ended September 30, 2005, cash used by operating activities was principally the result of our net loss for the period and increases in inventory and accounts receivable, partially offset by an increase in deferred revenue.

   Net cash used by operating activities was $36.7 million in 2005, $52.7 million in 2004 and $8.7 million in 2003. For 2005, net operating cash was used primarily by our net operating loss, decreases in accounts payable and payroll liabilities, and increases in accounts receivable and inventory, partially offset by an increase in deferred revenue. For 2004, net operating cash was used primarily by our net operating loss and an increase in inventory, offset in part by increases in deferred revenue and accounts payable and a decrease in accounts receivable. In 2003, net operating cash was used primarily by increases in accounts receivable, inventory and prepaid expenses and other assets, which was offset in part by an increase in deferred revenue.

   Net cash used by investing activities was $2.1 million for the nine months ended September 30, 2006, compared to net cash provided by investing activities of $41.9 million for the respective 2005 period. Net cash used by investing activities for the nine months ended September 30, 2006 consisted primarily of purchases of property and equipment. Net cash provided by investing activities for the same period in 2005 consisted of net sales and maturities of short-term investments of $34.3 million and a decrease in restricted cash of $11.4 million, partially offset by $3.8 million of purchases of property and equipment.

   Net cash provided by investing activities was $41.7 million in 2005, while net cash used in investing activities was $29.9 million in 2004 and $41.2 million in 2003. For the year ended December 31, 2005, net cash provided by investing activities consisted of the sale of short-term investments, partially offset by the purchases of short-term investments and equipment as well as

   Net cash provided by financing activities was $2.3 million for the nine months ended September 30, 2006, compared to net cash used by financing activities of $273,000 for the respective 2005 period. Cash provided by financing activities for both periods was primarily cash received from the exercise of stock options and the issuance of common stock through our employee stock purchase plan. In both periods, these proceeds were offset by line of credit issuance costs and principal payments on capital leases.

   Net cash used in financing activities was $137,000 in 2005, while net cash provided by financing activities was $84.2 million in 2004 and $65.1 million in 2003. For the year ended December 31, 2005, net cash used in financing activities consisted primarily of $755,000 paid for line of credit issuance costs and $731,000 for payments on capital leases, offset by $1.3 million in proceeds from the issuance of common stock through the employee stock purchase plan and exercise of stock options. The 2004 net cash provided by financing activities was primarily related to our Note offering in which we received net proceeds of $76.6 million. In 2004 we also received approximately $8.3 million through stock option and warrant exercises as well as through the issuance of common stock in connection with our employee stock purchase plan. The 2003 net cash provided by financing activities was primarily from our public offering, in which we received net proceeds of $42.5 million, and $27.0 million from stock option and warrant exercises and the issuance of common stock through the employee stock purchase plan, while we used $4.1 million to pay down a portion of our debt.

   Over the next twelve months, our significant cash requirements will relate to operational expenses, consisting primarily of personnel costs, costs of inventory and spare parts, outside engineering expenses, particularly as we continue development of our Cray XT4 and successor systems and internally fund a portion of the expenses on our Cascade project pursuant to the DARPA Phase III award, interest expense and acquisition of property and equipment. As of September 30, 2006, our remaining fiscal year 2006 capital budget for property and equipment was approximately $4.0 million. In addition, we lease certain equipment and facilities used in our operations under operating or capital leases in the normal course of business. The following table summarizes our contractual cash obligations as of September 30, 2006 (in thousands):

   We have $80.0 million in aggregate principal amount of outstanding Notes due in 2024. The Notes bear interest at an annual rate of 3.0%, or $2.4 million per year, and holders of the Notes may require us to purchase the Notes on December 1, 2009, December 1, 2014 and December 1, 2019 or upon the occurrence of certain events provided in the indenture governing the Notes. Additionally, we have a two-year revolving line of credit for up to $30.0 million, which expires in May 2007. No amounts were outstanding under this line as of September 30, 2006. As of the

   In our normal course of operations, we have development arrangements under which we engage outside engineering resources to work on our research and development projects. For the nine months ended September 30, 2006, we incurred $18.3 million for such arrangements.

   At any particular time, our cash position is affected by the timing of cash receipts for product sales, maintenance contracts, government co-funding for research and development activities and our payments for inventory, resulting in significant fluctuations in our cash balance from quarter-to-quarter and within a quarter. Our principal sources of liquidity are our cash and cash equivalents, operations and credit facility. Even assuming acceptances and payment for large new systems to be sold and the benefit from our 2004 and 2005 restructurings and other recent cost reduction efforts, our cash flow from operations may be negative for 2006 as a whole, including the fourth quarter of 2006, largely to support working capital requirements, although a wide range of results is possible. With this offering, and the near term expected cash flow from our DARPA Phase III award, we do not anticipate borrowing from our credit line and we expect our cash resources to be adequate for at least the next twelve months.

   We have been focusing on expense controls, negotiating sales contracts with advance partial payments where possible, implementing tighter purchasing and manufacturing processes and improving working capital management in order to maintain adequate levels of cash. Additionally, the adequacy of our cash resources is dependent on the amount and timing of government funding as well as our ability to sell our products, particularly the Cray XT4, BlackWidow and Cray XMT systems, with adequate margins. Beyond the next twelve months, the adequacy of our cash resources will largely depend on our success in re-establishing profitable operations and positive operating cash flows on a sustained basis. See “Risk Factors” above.

BUSINESS

Overview

   We design, develop, manufacture, market and service high performance computing (“HPC”) systems, commonly known as supercomputers. Our supercomputer systems provide capability, capacity and sustained performance far beyond typical server-based computer systems and address challenging scientific and engineering computing problems.

   We believe we are well positioned to meet the HPC market’s demanding needs by providing superior supercomputer systems with performance and cost advantages when sustained performance on challenging applications and total cost of ownership are taken into account. We differentiate ourselves from our competitors primarily by concentrating our research and development efforts on the processing, interconnect and software capabilities that enable our systems to scale — that is, to continue to increase performance as our systems grow in size. Purpose-built for the supercomputer market, our systems balance highly capable processors, highly scalable software and very high speed interconnect and communications capabilities.

   We focus our sales and marketing activities on government agencies, industrial companies and academic institutions that purchase high end HPC systems. We sell our products primarily through a direct sales force that operates throughout the United States and in Canada, Europe, Japan and Asia-Pacific. Our supercomputer systems are installed at more than 100 sites in over 20 countries.

   In 2005, our management changed significantly with a new chief executive officer and new leaders in technology, engineering, finance, marketing, operations and customer support. Under our new management team, we have expanded our worldwide customer base, refined our product roadmap, established a lower operating cost model and sharpened our focus on execution to meet customer expectations and improve our financial operating results.

   In early 2006 we announced our Adaptive Supercomputer vision to expand the concept of hybrid computing to a fully integrated view of both hardware and software supporting multiple processing technologies within a single, highly scalable system. We believe our recent $250 million award from DARPA under its HPCS program validates our Adaptive Supercomputer vision. This award will co-fund our Cascade development project to implement this vision.

Industry Background

   Since Seymour Cray introduced the Cray-1 system in 1976, supercomputers have contributed substantially to the advancement of knowledge and the quality of human life. Scientists and engineers typically require vast computing resources to address problems of major economic, scientific and strategic importance. Much of the development of new products and technologies, as well as improvements of existing products and technologies, will not be possible without the continued improvement of supercomputer computational speeds, interconnect technologies, scalable system software and overall performance.

   The overall server market is estimated by the International Data Corporation (“IDC”), in its reports entitled Worldwide and U.S. Server 2006-2010 Forecast and Worldwide Technical Computing Systems 2006-2010 Forecast, issued in April and May 2006, respectively, to have been $51.3 billion worldwide in 2005. According to these reports, the HPC market, which is a sub-sector of the overall server market, totaled $9.2 billion in 2005. We target the high end of the HPC market, which includes the capability segment, and a portion of the enterprise segment, as these segments are defined by IDC. We believe our total addressable market within these segments is approximately $1.5 billion in annual product sales.

   The capability segment is characterized by intensive research and development requirements and high barriers to entry necessary to deliver systems configured to solve the world’s largest and most demanding problems. Systems purchased in the enterprise segment primarily support high capacity requirements of many small and medium-sized technical applications running concurrently. Systems in these two market segments range in price from $1 million to $50 million or more.

   Vendors that compete in the highest end of the HPC market must commit significant resources to develop proprietary technologies and computing elements to meet the exacting needs of their customers. We believe that the technical requirements and high costs required to compete in this market are significant barriers to entry. Many of our potential competitors focus on the lower segments of the HPC market. These segments comprise a larger market that is increasingly competitive and in which it is difficult for vendors to add significant value due to the commoditization of these solutions.

   Supercomputer users are seeking answers to some of the world’s most complex problems in science and engineering. Addressing these challenges can require up to 1,000 times or more than the computing capability currently available with existing computer systems. Users require systems that provide powerful computing resources that are massively scalable, flexible and manageable at this scale, and that can deliver high levels of sustained performance.

   We believe there are three principal factors driving the demand for supercomputing power: first, the increasing need for advanced design and simulation capability in industry, government agencies and weather and climate centers; second, continuing concerns about national security issues, heightened by an emphasis on terrorism prevention; and, third, the recognized national interests of many countries to advance scientific research to enable innovations to better compete globally and achieve breakthroughs in new energy technologies, biological systems, nanotechnologies, particle physics and other natural phenomena.

   Design and simulation of new products before they are built are invaluable tools to improve time-to-market, product quality and differentiation for government, industrial and academic users. The need for supercomputers within government laboratories and agencies and industrial firms is driven by the increasingly complex application requirements of computer-aided engineering, full-systems analysis, material behavior in composite materials and real-time stress-strain behavior. Supercomputers are critical for increasingly refined simulations of both automotive and aeronautical performance dynamics. Weather forecasting and climate centers require supercomputers to process large volumes of data to produce more accurate short-term and medium-range forecasts and to further our understanding of the long-term impact of various pollutants on the environment and the effects of global warming.

   Governments have a wide range of ongoing and yet unmet security needs, ranging from burgeoning cryptanalysis and data mining requirements to rapid and accurate analysis of data from a diverse and growing number of disparate sources. In addition, governments constantly seek better simulation and modeling of missiles and other weapons systems and the maintenance and reliability of nuclear stockpiles. They also use supercomputers to simulate real world battlefield conditions rapidly and in increasing levels of detail.

   Competition between countries to acquire the best supercomputing technology to enhance their worldwide competitiveness has increased. The U.S. government and its various agencies have determined that it is in the best economic and security interest of the country to establish and maintain a leadership position in the development of supercomputing technologies. One such initiative is the DARPA High Productivity Computing Systems initiative, under which we have received funding since 2002 for our Cascade project. The DARPA program, announced in 2002, is designed to provide government support to develop breakthroughs in high productivity

   Despite the demand for increased supercomputing power, systems capable of exploiting high end opportunities have become less common. Today’s HPC market is replete with low bandwidth cluster systems that are often limited in performance beyond certain system size and capability. These systems loosely link together, or cluster, multiple commodity servers using widely available microprocessors by means of commercially available interconnect products.

   With standard interconnect components, low bandwidth cluster systems are not well balanced — they may have fast processors, but performance is severely limited by the rate at which data can be moved throughout the system, such as to and from memory, and between processors over the interconnection network. Because of the lack of specialized communication and software capabilities, these systems do not scale well — as these systems grow in size their full system and per processor efficiencies degrade significantly. Additionally, as these systems grow in size, they may become unreliable because they lack the necessary management software and built-in hardware redundancies to minimize disruptions.

   Low bandwidth cluster systems may offer higher theoretical peak performance, for equivalent cost, than do our systems, but often lack in sustained performance when running real applications at scale. Theoretical peak performance is the highest theoretical possible speed at which a computer system could, but never does, operate; this measure is obtained simply by multiplying the number of processors by their peak rated speed and the number of floating point operations per cycle it can compute, assuming zero communications bottlenecks or system inefficiencies. Sustained performance, always lower than peak, is the actual speed at which a supercomputer system runs an application program. The sustained performance of low bandwidth cluster systems on complex applications frequently is a small fraction, often less than 5% to 10%, of their theoretical peak performance — as these systems become larger, their efficiency declines even further, sometimes below 1% for the most challenging applications at scale.

   The recent introduction of dual-core processors, and planned quad-core and multi-core processors, which incorporate more than one processing core on the same integrated circuit, will further stress the capabilities of low bandwidth cluster systems, resulting in decreased per processor utilization due to the absence of balanced network and communication capabilities in such systems. Multi-core processors will also increase the power and cooling requirements for these systems, making packaging an increasingly critical element.

   Given these limitations, low bandwidth cluster systems are better suited for applications that can be partitioned easily into discrete tasks that do not need to communicate often with each other, such as small problems and larger problems lacking communications complexity that make up the majority of the midrange and low end of the HPC market. The effectiveness of low bandwidth cluster systems in our target market, the high end of HPC, is limited today and, we believe, will be increasingly limited in the future.

The Cray Solution

   We have concentrated our product roadmap on building balanced systems that are purpose-built for supercomputer users. These systems address the critical computing resource challenges HPC users face today: achieving massive scaling to tens of thousands of processors, ease of use, and very high levels of sustained performance on real applications. We do this by designing supercomputers that combine highly capable processors, whether developed by us or by others,

   Our supercomputers utilize components and technologies designed to support the demanding requirements of high end HPC users. In contrast, low bandwidth cluster system vendors use processors, interconnects and software designed to meet the requirements of the significantly larger general purpose server market and then attempt to leverage these commercially-oriented products into the HPC market. An important benefit of our purpose-built approach is significantly higher sustained performance on real applications, with actual application performance improvements on the order of 1.5 to 10 times that of our competitors. With our supercomputers, HPC users are able to focus on their primary objectives: advancing science, industry and discovery and improving national security.

   Our supercomputer systems offer several additional benefits:

   We expect the emergence of multi-core processors to be advantageous to us, complementing our technical strengths in networking, scaling software, and cooling and power management technologies. Additional cores will amplify the scaling issues that customers face today by putting increased stress on all aspects of the system. Our balanced approach to system design will likely become increasingly critical in enabling customers to take advantage of the benefits of multi-core processing.

Our Strategy

   Our goal is to become the leading provider of supercomputers in the markets that we target. Key elements of our strategy include:

   Gain Share in Our Core HPC Market. We intend to leverage our strong product portfolio, product roadmap and brand recognition in the high end of the HPC market to gain market share. We believe that most of our competitors are primarily focused on the lower end of the HPC market where low-bandwidth cluster systems dominate. We plan to remain focused on the capability and enterprise segments of the HPC market.

   Maintain Focus on Execution and Profitability. We are committed to achieving sustained profitability on an annual basis. We intend to continue to refine our product roadmap, converge our technologies and development processes, improve our ability to deliver high quality products on time and on budget and continue our commitment to financial discipline.

   Extend Technology Leadership. We are an innovation driven company in a technology driven market. We plan to maintain a technology leadership position by investing in research and development and partnering with key customers with interests aligned strongly with ours. We will rely in part on government funding for our research and development efforts. We intend to execute on our product roadmap and implement our Adaptive Supercomputing vision to realize the concept of supporting multiple processing technologies within a single, highly scalable Linux-based system.

   Expand Total Addressable Market. Over time, we intend to leverage our technologies, customer base and Cray brand in new segments and expand our addressable market. We believe we have the opportunity to compete in a broader portion of the HPC market as well as selective markets outside of HPC.

Our Target Market and Customers

   Our supercomputer systems are installed at more than 100 sites in over 20 countries. Our target markets for 2006 and beyond principally include the national security, scientific research, earth sciences, and computer-aided engineering, consisting primarily of automotive, aerospace and manufacturing companies, markets. In certain of our targeted markets, such as the national security and scientific research markets, customers have their own application programs and are accustomed to using new, less proven systems. Other target customers, such as automotive and aerospace firms and some governmental agencies, require third-party application programs developed by independent software vendors.

   National Security. Classified work in government agencies has represented an important customer market for us over many years. Certain governmental departments continue to provide funding support for our research and development efforts to meet their objectives. We expect long-term spending on national security and defense to increase. Current and target customers for our products include a number of Department of Defense-related classified customers, the National Nuclear Security Administration of the Department of Energy, and certain foreign counterparts.

   Scientific Research. Scientific research includes both unclassified governmental and academic research laboratories and centers. The Department of Defense, through its High Performance Computing Modernization Program, funds a number of research organizations that are target customers. The Office of Science in the Department of Energy and its laboratories are key target customers, as are the National Science Foundation and the National Aeronautics and Space Administration, and related agencies around the world.

   Earth Sciences. Weather forecasting and climate modeling applications require increasing speed and larger volumes of data. Forecasting models and climate applications have grown increasingly complex with an ever-increasing number of interactive variables, making improved supercomputing capabilities increasingly critical. We have a number of customers doing weather and climate applications and believe that the Cray X1E system installed at the Korea Meteorological Administration is currently the most powerful operational weather forecasting system in the world.

   Computer-Aided Engineering. Supercomputers are used to design lighter, safer and more durable vehicles, as well as to study wind noise and airflow around the vehicle, to improve airplane flight characteristics and in many other computer-aided engineering applications in order to improve time-to-market and product quality. We currently have customers in each of the automotive, aerospace and manufacturing areas.

   Agencies of the U.S. government, directly and indirectly through system integrators and other resellers, accounted for approximately 55% of our 2005 revenue, 74% of our 2004 revenue and 74% of our 2003 revenue. Our largest customer in 2005 was Oak Ridge National Laboratory with 18% of our revenue, in 2004 was Sandia National Laboratories through the Red Storm project with 27% of our revenue and in 2003 Oak Ridge National Laboratory had 11% of our total revenue. International customers accounted for 32% of our total revenue in 2005, up from 17% in 2004 and 18% in 2003, showing our strong growth in international markets.

Recent Customer Contract Wins

   We have had significant recent customer contract wins which we believe are indicative of the value that we bring to our customers. The following represent announced 2006 contract wins:

   Sandia National Laboratories, part of the U.S. Department of Energy’s National Nuclear Security Administration, upgraded its Red Storm supercomputer from 40 teraflops to 125 teraflops (40 to 125 trillion floating point operations per second), making it the second most powerful supercomputer in the world, as announced in November 2006 on the “Top 500” list. The Red Storm system, co-designed by Sandia and ourselves and built by us, is the basis for the Cray XT3 system and is a key component of Sandia’s ongoing mission of putting advanced technology to work in the pursuit of global safety, peace and freedom. This achievement was announced in November 2006.

   CSC Finland (“CSC”), the information technology center for science in Finland, will acquire a Cray XT4 system delivering over 70 teraflops of compute power. The system will be installed in stages, beginning in late 2006 and continuing through 2008. CSC provides information technology infrastructure, skills and specialist services for universities, polytechnic colleges, research institutions and companies across Finland, and collaborates with various research institutions worldwide. The Cray XT4 system will be used for research in areas such as physics, chemistry, nanotechnology, linguistics, bioscience, applied mathematics and engineering. We were told that CSC selected the Cray XT4 system after an extensive acquisition process that involved surveying 35 different research groups, closely analyzing the available technologies and benchmarking competing systems. The contract was announced in October 2006.

   National Energy Research Scientific Computing Center (“NERSC”), a laboratory of the U.S. Department of Energy’s Office of Science, awarded us a $52 million contract for products and services to deliver our Cray XT4 system in 2007, with options to provide future upgrades that could quadruple the performance of the system and boost performance to one petaflops and beyond. NERSC is one of the largest scientific computing facilities in the world devoted to providing computational resources and expertise for a broad base of unclassified research. We were informed that our proposal was selected because of its price/performance and overall effectiveness, as determined by NERSC’s comprehensive evaluation criteria of more than 40 measures. This contract win was announced in August 2006.

   Oak Ridge National Laboratory (“ORNL”), the largest laboratory of the Department of Energy’s Office of Science and its current “Leadership Computing” center, awarded us a $200 million contract for products and services to be provided in progressive upgrades to ORNL’s existing Cray XT3 supercomputer and future systems being developed under our Cray XT4 and Baker programs. The Baker system, planned for delivery toward the end of 2008 or early 2009, is expected to provide peak performance of one petaflops (1,000 trillion floating point operations per second). ORNL is an international leader in research areas that include neutron science, new energy sources, biological systems, nanoscale materials science and national security. This contract win was announced in June 2006.

   Swiss National Supercomputing Center (“CSCS”) awarded us a contract for products and services to expand Europe’s first Cray XT3 supercomputer from 5.7 teraflops to more than 8.6 teraflops. Large scientific research projects targeted for the Cray XT3 system at CSCS include analysis of human bone structure, the environmental effects of aerosols and pollutants, anticancer drugs, nanoelectronics, molecular switches and global climate model. This contract win was announced in May 2006.

   The University of Western Australia awarded us a contract to purchase a Cray XT3 system and related services as part of the Western Australia Supercomputing program. The Cray XT3 system is being used for major large-scale computational studies and simulations in the areas of

   The Institute for Plasma Research (“IPR”), headquartered in Gujarat, India, awarded us a contract for a Cray X1E system and services. IPR will use the Cray X1E system to conduct studies in plasma physics with an emphasis on the physics of magnetically confined hot plasmas and nonlinear plasma phenomena. The Cray X1E system provides IPR with the processing power and scalability that it needs to design effective research devices and conduct fruitful experiments. This contract win was announced in March 2006.

   AWE Plc awarded us a £20 million contract for product and services to deliver a dual-core Cray XT3 system with peak performance of over 40 teraflops in 2006. The new system will expand AWE’s current computing power by almost 30 times, as measured on AWE’s rigorous benchmark tests, and will enable AWE to explore even more complex mathematical models in three dimensions and to make advances on a range of scientific fronts, including weapon physics, materials science and engineering. AWE, a United Kingdom governmental agency, is one of the largest high technology research, design development and production facilities in that country. This contract win was announced in January 2006.

Current Products and Products in Development

   Our supercomputers provide capability, capacity and sustained performance far beyond typical server-based computer systems, allowing users to address challenging scientific and engineering computing problems. Purpose-built for the supercomputing market, our systems balance highly capable processors, highly scalable software and very high speed interconnect and communications capabilities. We plan to utilize increasingly common infrastructure, components and system software pursuant to our Adaptive Supercomputing vision. Our goal is to bring new products and/or major enhancements to market every 12 to 24 months.

   The following table lists our current products and products in development by internal code names:

   Cray XT3 System. The Cray XT3 system uses AMD single-core and dual-core Opteron processors connected via our proprietary high bandwidth interconnect network. It incorporates a massively parallel tailored operating system and a standards-based programming environment designed to deliver very high sustained application performance in configurations from 100 to over 30,000 processors. The Cray XT3 system, based on the Red Storm architecture we co-developed with Sandia National Laboratories, features a tightly integrated management system to provide high reliability and enable full-system applications to run to completion. We began shipments of early versions of the Cray XT3 system in the fourth quarter of 2004, with full production ramp in the first half of 2005. Our selling focus for the Cray XT3 system covers a range of peak performance from one to over 100 teraflops.

   Cray XT4 System. Our Cray XT4 system combines the capabilities of our Cray XT3 system and many software features of our Cray XD1 system to provide a next generation massively parallel processor supercomputer system. Our Cray XT4 system uses dual-core AMD Opteron processors, which are field-upgradeable to quad-core, running a lightweight Linux operating system and connected to our proprietary second generation high speed network. The Cray XT4 system is highly scalable and is designed to provide significant improvements in peak and sustained performance.

   Cray X1E System. In late 2002 we completed hardware development of the Cray X1 system, which incorporates in its design both vector and massively parallel processing capabilities. We commenced delivering production systems late in the fourth quarter of 2002 and had full production ramp in 2003. The Cray X1E system, first shipped in December 2004, nearly triples the peak performance of the Cray X1 system per cabinet and features one of the world’s most

   Cray XD1 System. The Cray XD1 system, designed for the midrange HPC market, uses a Linux-based operating system in concert with our automated management infrastructure and provides the opportunity to accelerate application performance through the use of field programmable gate arrays. We plan to combine the capabilities of the Cray XD1 and Cray XT3 systems into our Baker system in development.

Products in Development

   BlackWidow. Our BlackWidow program is directed at developing our next generation vector-based supercomputer as a successor to our Cray X1E system. The BlackWidow system is designed to provide major improvements in single thread scalar performance and overall price performance as measured on both peak and sustained bases. The BlackWidow system will be tightly coupled with our Cray XT3 and Cray XT4 systems so that the user sees a unified environment and file system across both products, representing an important step in our program towards providing a heterogeneous computing environment. The BlackWidow program is co-funded by the U.S. government.

   Cray XMT. Our Cray XMT program is directed at developing a third generation multithreaded supercomputer, which offers global shared memory and high latency tolerance, with 128 threads per processor. The Cray XMT system will utilize our Cray XT3 infrastructure and is a significant step towards implementing our Adaptive Supercomputing vision. The Cray XMT program is co-funded by the U.S. government.

   Baker. Our Baker program is directed at creating the successor to our Cray XT4 system and to extend our leadership position in massively parallel computing. The Baker system will utilize a new highly configurable interconnect system that combines the interconnect technologies of the Cray XT3 and Cray XD1 systems and next generation quad-core and multi-core AMD Opteron processors in a more densely packaged air and/or liquid-cooled cabinet. The Baker system is expected to provide beyond one petaflops peak performance. Our recent contract with Oak Ridge National Laboratory was the first announced contract for a petaflops performance system and is based on our Baker system.

Our Adaptive Supercomputing Vision and Cascade Program

   Our Adaptive Supercomputing vision supports the anticipated future needs of HPC customers. With Adaptive Supercomputing, we expect to expand the concept of heterogeneous computing to a fully integrated view of both hardware and software supporting multiple processing technologies within a single, highly scalable system. Our plan is to increasingly integrate these processing technologies into a single Linux-based platform. We expect to include powerful compilers and related software that will analyze and match application codes to the most appropriate processing elements — we expect this capability will enable programmers to write code in a more natural way. We believe the recent DARPA $250 million award to us validates this vision, which was the center of our DARPA HPCS Phase III proposal.

   Our Adaptive Supercomputing vision incorporates many of our technical strengths — system scalability, multiple processing technologies, including proprietary processors, and high bandwidth networks — into a single system that we believe will make supercomputing capabilities accessible to a larger set of end-users.

   Our Cascade development program implements our Adaptive Supercomputing vision by easing the development of parallel software codes, supporting global address space models which exploit shared memory and providing for new high productivity languages. We plan to develop

Technology

   Our leadership in supercomputing is dependent upon the successful development and timely introduction of new products. We focus our research and development activities on designing system architecture, hardware and software necessary to implement our product roadmap.

   We believe we are the only company in the world with significant demonstrated expertise in four primary processor technologies: vector processing, massively parallel processing, multithreading and co-processing with field programmable gate arrays.

   Cray Research pioneered the use of vector systems. These systems traditionally have a moderate number of very fast custom processors utilizing shared memory. Vector processing is the computation of a vector or string of numbers with a single operation. This technology has proven to be highly effective for many scientific and engineering applications in areas such as weather forecasting, cryptanalysis and computational fluid dynamics. Vector processing is the basis for our existing Cray X1E system and our successor BlackWidow product.

   Massively parallel processing architectures typically link hundreds or thousands of commodity processors together in a single system. These systems are best suited for large computing problems that can be segmented into many parts and distributed across a large number of processors. We focus on building systems with highly scalable architectures using high bandwidth interconnect networks. The Cray XT3, Cray XT4 and the Cray XD1 systems, and the Baker successor system, are based on this architecture.

   Multithreading is designed to provide latency tolerance by supporting a large number of executable threads per processor, and quickly switching to another thread when a thread stalls waiting for data to be computed or to return from memory. These systems are particularly effective for irregular access to large data sets and graph-based algorithms. We are currently developing a third generation multithreading system as part of our Cray XMT project.

   Field programmable gate arrays can be reconfigured or reprogrammed to implement specific functionality more suitably and more efficiently than on a general-purpose processor. The Cray XD1 system introduced the concept of reconfigurable computing with field programmable gate arrays to our product portfolio, and we have a roadmap which will bring reconfigurable computing to our Cray XT3 and Cray XT4 systems.

   We have extensive experience in designing hardware components of HPC systems — processors, memory controllers, interconnect systems, I/ O subsystems and cooling, power, and packaging infrastructures — and integrating them into a single system. Our hardware research and development experience includes:

   Our hardware engineers are located primarily in our Chippewa Falls, Wisconsin, and Seattle, Washington, offices.

   We have extensive experience in designing and developing software for HPC systems. This includes the operating system, the hardware supervisory system and the programming environment. Over time we plan to transition to a common system software and a common programming environment across all of our platforms, an important aspect of our Adaptive Software vision. Our software research and development experience includes:

   We purchase or license software technologies from third parties when necessary to provide appropriate support to our customers, while focusing our own resources where we believe we add the highest value.

   Our software personnel are located principally in our Mendota Heights, Minnesota, and Seattle, Washington, offices.

Services

   Our worldwide service organization provides us with a competitive advantage and a predictable flow of revenue and cash. Support services are important to our customers, and we generally locate our support personnel at customer locations. In recent years, annual service revenue has ranged from approximately one-quarter to one-third of total revenue. Our support services include facility analysis, system installation, application porting, tuning and support, hardware maintenance and operating system support.

   Support services are provided under separate maintenance contracts with our customers. These contracts generally provide for support services on an annual basis, although some cover multiple years. While most customers pay for support monthly, others pay on a quarterly or annual basis. Customers may select levels of support and response times, ranging from parts only to 24 x 7 coverage with two-hour response.

   Our professional services, which include product integration, advanced computer training, project management services, site engineering, application analyst support and customer hardware and software engineering, are provided on a project-by-project basis.

   We have approximately 120 employees in our service organization located at or near customer sites globally and approximately 50 employees in our central service organization located in Chippewa Falls, Wisconsin, and Mendota Heights, Minnesota.

Sales and Marketing

   We focus our sales and marketing activities on government agencies, industrial companies and academic institutions that purchase HPC systems. We sell our products primarily through a seasoned supercomputing direct sales force that operates throughout the United States and in Canada, Europe, Japan and Asia-Pacific. We serve smaller foreign markets through sales representatives and resellers. About half of our sales force is located in the United States and Canada, with the rest overseas.

   A majority of our sales are driven by a formal request for proposal process for HPC systems. We utilize pre-sales technical experts to develop technical proposals that meet the customer requirements and benchmarking teams to demonstrate the advantages of our particular supercomputing products being proposed. For a majority of sales opportunities, the terms of our proposals, including system size, options, pricing and other commitments, are individually reviewed and approved by our senior executives. While we often tailor our supercomputer solutions for each customer, there is substantial commonality in the underlying components and systems, allowing us to mitigate potential impacts on manufacturing and procurement operations.

   As government agencies and government funded scientific research institutions comprise a large portion of our customer base, our government programs office is an integral part of our overall sales and marketing strategy. Our government programs staff actively manages our relationship with U.S. government agencies and Congress.

   Our marketing staff is responsible for product marketing, marketing communications and business development. Product marketing bridges our research and development organization and our sales staff to help ensure that our products meet the demands and requirements of our key customers and a broader set of prospects. Marketing communications focus on our overall brand messaging, press releases, conferences, trade shows and marketing campaigns. Business development focus on providing products and services to specific customer sets, such as earth sciences and computer-aided engineering.

Intellectual Property

   We attempt to protect our trade secrets and other proprietary rights through formal agreements with our employees, customers, suppliers and consultants, and through patent protection. Although we intend to protect our rights vigorously, there can be no assurance that our contractual and other security arrangements will be successful. There can be no assurance that such arrangements will not be terminated or that we will be able to enter into similar arrangements on favorable terms if required in the future. In addition, if such agreements were breached, there can be no assurance that we would have adequate remedies for any breach.

   Our general policy is to seek patent protection for those inventions and improvements likely to be incorporated into our products and services or to give us a competitive advantage. We have a number of patents and pending patent applications relating to our hardware and software technologies. While we believe our patents and applications have value, no single patent or group of patents is in itself essential to us as a whole or to any of our key products. Any of our proprietary rights could be challenged, invalidated or circumvented and may not provide significant competitive advantage.

   We license certain patents and other intellectual property from Silicon Graphics as part of our acquisition of the Cray Research operations. These licenses contain restrictions on our use of the underlying technology, generally limiting the use to historic Cray products, vector processor computers and the Cray X1/ X1E system.

   There can be no assurance that the steps we take will be adequate to protect or prevent the misappropriation of our intellectual property. We may infringe or be subject to claims that we infringe the intellectual property rights of others. Litigation may be necessary in the future to enforce patents we obtain, and to protect copyrights, trademarks, trade secrets and know-how we own, or to defend infringement claims from others. Such litigation could result in substantial expense to us and a diversion of our efforts.

Manufacturing and Procurement

   We subcontract the manufacture of substantially all of the hardware components for all of our products, including integrated circuits, printed circuit boards, connectors, cables, power supplies and memory parts, on a sole or limited source basis to third-party suppliers. We use contract manufacturers to assemble our components for all of our systems. Our manufacturing strategy is on build-to-order systems, focusing on obtaining competitive assembly and component costs and concentrating on the final assembly, test and quality assurance stages. This strategy allows us to avoid the large capital commitment and overhead associated with establishing full-scale manufacturing facilities and to maintain the flexibility to adopt new technologies as they become available without the risk of equipment obsolescence, provide near real-time configuration changes to exploit faster and/or less expensive technologies, and provide a higher level of large scale system quality. We perform final system integration, testing and check out of our hardware systems. Our manufacturing personnel are located primarily in Chippewa Falls, Wisconsin.

   Our systems incorporate some components that are available from single or limited sources, often containing our proprietary designs. Such components include integrated circuits, interconnect systems and certain memory devices. Prior to development of a particular product, proprietary components are competitively bid to a shortlist of technology partners. The technology partner that provides the best solution for the component is generally awarded the contract for the life of the component. Once we have engaged a technology partner, changing our product designs to utilize another supplier’s integrated circuits can be a costly and time-consuming process. We also have sole or limited sources for less critical components, such as peripherals, power supplies, cooling and chassis hardware. We obtain key integrated circuits from IBM for our Cray X1E, Cray XT3 and Cray XT4 systems, from Texas Instruments for our BlackWidow project and from Taiwan Semiconductor Manufacturing Company for our Cray XMT project and microprocessors from AMD for our Cray XT3, Cray XT4, Cray XD1 and successor systems. Our procurements from these vendors are primarily through purchase orders. We have chosen to deal with sole sources in specific cases due to the availability of specific technologies, economic advantages and other factors. Reliance on single or limited source vendors involves several risks, including the possibility of shortages of key components, long lead times, reduced control over delivery schedules and changes in direction by vendors. See “Risk Factors — Our reliance on third-party suppliers poses significant risks to our business and prospects” above.

Competition

   The HPC market is very competitive. Many of our competitors are established companies well known in the HPC market, including IBM, NEC, Hewlett-Packard, Silicon Graphics, Dell, Hitachi, Bull and Sun Microsystems. Most of these competitors have substantially greater research, engineering, manufacturing, marketing and financial resources than we do.

   We also compete with systems builders and resellers of systems that are constructed from commodity components using microprocessors manufactured by Intel, AMD, IBM and others. These competitors include the previously named companies as well as smaller firms that benefit from the low research and development costs needed to assemble systems from commercially available commodity products. These companies have capitalized on developments in parallel processing and increased computer performance in commodity-based networking and cluster systems. While these companies’ products are more limited in applicability and scalability, they have achieved growing market acceptance as they offer significant peak/price performance on