Item 1. Business

Rambus Inc. ("Rambus" or the "Company") designs, develops, licenses and markets high-speed chip-to-chip interface technology to enhance the performance and cost-effectiveness of computers, consumer electronics and other electronic systems. The Company licenses semiconductor companies to manufacture and sell memory and logic ICs incorporating Rambus interface technology and markets its solution to systems companies to encourage them to design Rambus interface technology into their products. The Company's technology cost-effectively increases the data transfer rate, or "memory bandwidth," allowing semiconductor memory devices to keep pace with faster generations of processors and controllers and thus supports the accelerating data transfer requirements of multimedia and other high-bandwidth applications.

Rambus was incorporated in California in March 1990 and reincorporated in Delaware in March 1997.

Background

The performance of a computer or other electronic system is typically constrained by the speed of its slowest element. In the past, that element was the logic IC that controlled the system's specific functions and performed calculations-the microprocessor. In recent years, however, new generations of microprocessors and controllers have become substantially faster and more powerful, and increasingly the bottleneck in system performance is becoming the component that stores the instructions and data needed by the microprocessors and controllers-the DRAM.

Since 1980, the typical operating frequency of Intel and other mainstream microprocessors has increased approximately 40 times from 5 MHz (million cycles per second) to over 200 MHz. During this same period, the typical operating frequency of a standard DRAM (known as a "page mode" DRAM) has increased by approximately five times. Even the evolutionary improvements to DRAM technology, including the FPM (fast-page mode) DRAM, the EDO (extended-data-out) DRAM, the SDRAM (synchronous DRAM) and the SGRAM (a version of the SDRAM for graphics applications), have only been able to provide an improvement of up to approximately ten times. This growing disparity between the frequency of microprocessors and DRAMs is termed the "Performance Gap."

While microprocessors have undergone both manufacturing and architectural improvements, significant innovations for DRAMs have generally only occurred on the manufacturing side. DRAM manufacturers have been successful in increasing DRAM "density," or storage capacity, from roughly 1 Kbit (thousand bits) to 64 Mbits (million bits) per chip, thereby reducing the number of DRAMs required for a given amount of memory. However, corresponding architectural improvements necessary to increase DRAM data transfer rates to keep pace with increasing microprocessor speeds have not occurred.

Rambus Technology

Rambus has created a revolutionary chip-to-chip interface architecture, which allows data to be transferred through a simplified bus at significantly higher frequencies than permitted by conventional technologies. Rambus has focused

the application of its interface technology on the Performance Gap and licenses its interface technology to memory and logic semiconductor manufacturers, which incorporate this interface technology into their IC designs to supply systems companies with Rambus ICs. The key elements of the Rambus interface are Rambus-based DRAMs ("RDRAMs"), Rambus ASIC cells ("RACs") and the interconnecting circuitry known as the "Rambus Channel." While Rambus technology can be used to address a wide variety of chip-to-chip data transfer requirements, the largest immediate application is to connect logic circuits to memory in home video games, PCs, workstations and other electronic systems.

Rambus interface technology currently allows data transfers of up to 600 megabytes per second between a logic IC and DRAMs by transferring data at a frequency of 600 MHz over a byte-wide bus. System performance can be further enhanced by applying Rambus interface technology to multiple channels on a logic
IC. For example, a Rambus-based logic IC can utilize four channels to achieve data transfer of up to 2.4 gigabytes per second. In addition, Rambus is working with Intel and others on the definition of an extension to the Rambus interface technology for 64 Mbit generation RDRAMs, scheduled for introduction in late 1998 and called "Direct Rambus" technology, which will be further optimized for PC main memory applications. This technology is scheduled to increase the operating frequency of RDRAMs to 800 MHz and double the bus width to allow data transfers of up to 1.6 gigabytes per second. There can be no assurance that the dates or performance criteria referred to in the preceeding forward-looking statements will be obtained. See "Factors Affecting Future Results - Future Dependence upon PC Main Memory Market Segment and Intel" and "- Rapid Technological Change; Reliance on Fundamental Technology; Importance of Timely New Product Development."

Target Markets and Applications

The high-speed interface technology Rambus has developed is applicable to data transfer between most semiconductor chips. The Company has chosen to concentrate the application of its technology on the interface between logic ICs and memory devices because of the acute performance needs and the relevant market sizes. While Rambus interface technology is useful in providing increased memory bandwidth in any electronic system, the Company believes that the systems which will best utilize the high bandwidth provided by current Rambus technology are the relatively high-volume, low-cost systems in which the cost of the memory subsystems represents a significant portion of the selling price. To date, the principal applications for the Company's technology have been in the consumer multimedia, PC multimedia and workstation multimedia markets. These areas accounted for the sale of approximately $13,000, $11 million and $447 million of Rambus ICs by Rambus licensees in calendar 1994, 1995 and 1996, respectively.

In November 1996, Rambus entered into a development and license contract with Intel. The contract provides for the parties to cooperate in the development of a specification for Direct Rambus next-generation 64 Mbit RDRAMs, which will be targeted at the PC main memory market segment. The contract also calls for Intel to use reasonable best efforts to develop a PC main memory controller designed for use with these RDRAMs. The Company believes that Direct Rambus technology will offer superior bandwidth compared to other solutions for PC main memory applications. However, these RDRAMs and the Intel memory controllers are not scheduled for mass production until 1999, and there can be no assurance that such devices will be successfully developed or that, if developed, will be successful in penetrating the market segment for PC main memory.

The Company believes that its technology, which enables high memory bandwidth at low cost, is well suited to a broad range of other applications. Other Rambus-based applications currently being developed include multifunction peripheral controllers for use in combination fax/copier/scanner/laser printer applications and networking equipment such as high-speed ethernet switches. There can be no assurance that such devices will be designed incorporating Rambus interface technology or that sales of such devices will be meaningful.

Rambus Licensees

Rambus licenses its technology on a nonexclusive and worldwide basis to semiconductor manufacturers who sell Rambus ICs to systems companies which have adopted Rambus technology. An important element of the Company's strategy is to license its technology broadly in order to establish Rambus interface technology as a standard and to provide systems companies with sources for Rambus ICs from established semiconductor companies. Rambus provides licenses to both DRAM manufacturers and logic IC manufacturers, who can license Rambus interface technology for use in producing RDRAMs and/or logic ICs containing RACs. At September 30, 1997, Rambus had a total of 21 licensees.

Rambus licensees include thirteen DRAM manufacturers which collectively accounted for over 90% of worldwide DRAM sales in calendar 1996: Fujitsu, Hitachi, Hyundai Electronics, IBM, LG Semicon, Micron Technology, Mitsubishi, NEC, Oki Electric Industry, Samsung Electronics, Siemens, Texas Instruments and Toshiba. At September 30, 1997, four of these licensees were in production of RDRAMs and nearly all were in development of next-generation Direct RDRAMs. Rambus logic licensees include Chromatic Research, Cirrus Logic, Fujitsu, IBM, Intel, LG Semicon, LSI Logic, NEC, SGS-THOMSON, Texas Instruments and Toshiba. At September 30, 1997, four of these licensees were in production of logic ICs which include RACs.

Rambus Business Model and Strategy

In order to establish Rambus interface technology as an industry standard, the Company has adopted an innovative business model in which it neither manufactures nor sells semiconductors incorporating the Company's technology. The Company licenses its technology on a nonexclusive and worldwide basis to semiconductor companies which manufacture and sell RDRAMs and logic ICs containing RACs to systems companies which have adopted Rambus technology. Systems companies are not required to obtain a Rambus license to incorporate Rambus ICs into their products. However, an important part of the Company's strategy is to maintain close ties to these systems companies to encourage the adoption of Rambus technology.

The Rambus business model and strategy are designed to promote Rambus as an industry standard, target leading systems companies in markets that the Company believes represent the greatest potential for Rambus IC sales, provide systems companies with multiple sources for RDRAMs, share research and development efforts with licensees, maintain technology leadership, pursue a system-level approach and generate revenue through a combination of contract fees and royalties.

Contract fees have provided the majority of the capital needed to date by the Company to develop its fundamental technology, and the Company believes that its business model is well suited to continue funding future development. However, there is no assurance that the Company's current partner licensees will generate revenue, or that the Company will be able to add new license contracts in the future, at levels sufficient to provide significant funding for further development activities.

Royalties, which are generally a percentage of the revenues received by licensees on their sales of Rambus ICs, are normally payable by a Rambus licensee on sales occurring during the life of the Rambus patents being licensed. For a typical systems application of Rambus technology, the Company receives royalties from the sale of both logic ICs containing RACs and RDRAMs as they are shipped by Rambus licensees. Royalty rates range up to a maximum of approximately 2.5% for RDRAMs and a maximum of approximately 5% for logic ICs, and in some cases may decline based on the passage of time or on the total volume of Rambus ICs shipped. The exact rate and structure of a royalty arrangement with a particular licensee depend on a number of factors, including the amount of the license fee to be paid by the licensee and the marketing and engineering commitment made by the licensee.

Design and Manufacturing

Rambus interface technology has been developed to allow semiconductor companies to use familiar, widely-available design tools and conventional techniques when designing their Rambus-enabled chips. A new Rambus licensee receives an implementation package from the Company which contains all the information needed to develop a Rambus IC in the licensee's process. There are separate implementation packages for RDRAMs and for RACs. An implementation package includes a specification, a generalized circuit layout database for the particular version of the RDRAM or RAC which the licensee intends to develop, test parameter software and, for RDRAMs, a DRAM core interface specification. Many licensees have contracted to have Rambus produce the specific implementation required to optimize the generalized circuit layout for the licensee's manufacturing process. In such cases, the licensee provides specific design rules and transistor models which Rambus designers use to integrate RDRAM or RAC circuits into the licensee's process. However, Rambus anticipates that as licensees become more familiar with the Rambus technology, they will be able to do more of the implementation work without Rambus' assistance.

Rambus has developed its technology to be manufacturable using familiar, industry-standard CMOS semiconductor processes. For this reason the Company believes that the wafer fabrication yields of RDRAMs and logic products containing RACs are consistent with those for similar products in the same manufacturing facility. However, because of the extra Rambus interface circuitry, an RDRAM chip is somewhat larger than a standard DRAM. Therefore, a

manufacturer will generally produce fewer RDRAMs than standard DRAMs for a given wafer size and an RDRAM chip will be somewhat more expensive than the standard version. Rambus believes that this die size premium is about 10% to 20% for the current 16 Mbit generation, but will be reduced to a target of less than 10% for 64 Mbit Direct RDRAMs. In addition, RDRAM manufacturers are responsible for their own manufacturing processes and Rambus has no role in the manufacture of RDRAMs. For example, Rambus has no influence on decisions in regard to any process changes or on whether or when to "shrink" or otherwise change a design to reduce the cost of the chips.

Current implementations of RDRAMs and Rambus logic ICs can be packaged in widely available, inexpensive packaging. System companies connect RDRAMs to Rambus logic ICs using normal printed circuit board ("PCB") materials and manufacturing techniques. System companies are provided with detailed specifications from Rambus on circuit board layout and construction. Circuit boards can be fabricated and assembled using standard PCB techniques and equipment.

Research and Development

The ability of the Company to compete in the future will be substantially dependent on its ability to advance its interface technology in order to meet changing market needs. To this end, Company engineers are involved in developing new versions of the Rambus interface technology which will allow chip-to-chip data transfer at higher speeds as well as provide other improvements. The Company has assembled a team of highly skilled engineers whose activities are focused on further development of Rambus interface technology as well as adaptation of current technology to specific licensees' processes. Because of the complexity of these activities, the design and development process at Rambus is a multi-disciplinary effort requiring expertise in computer architecture, digital and analog circuit design and layout, DRAM and logic semiconductor process characteristics, packaging, PCB routing and high-speed testing techniques.

As of September 30, 1997, Rambus had 99 employees in its design engineering departments. Approximately two thirds of these employees have advanced technical degrees. In fiscal 1995, 1996 and 1997, research and development expenses were approximately $3.1 million, $5.2 million and $9.8 million, respectively. In addition, because the Company's license agreements often call for engineering support by Rambus, a substantial portion of the Company's total engineering costs has been allocated to cost of contract revenues, even though these engineering efforts have direct applicability to Rambus' technology development. The Company expects that it will continue to invest substantial funds in research and development activities. There can be no assurance that new versions of the Rambus interface technology can be developed and introduced by the Company's licensees in a timely fashion or that such new technology will be accepted by the market. Moreover, the end markets for the Company's technology, particularly the home video game and PC markets, are subject to rapid technological change and there can be no assurance that as such markets change the Company's interface technology will remain current and suitable.

Competition

The semiconductor industry is intensely competitive and has been characterized by price erosion, rapid technological change, short product life cycles, cyclical market patterns and increasing foreign and domestic competition. All major DRAM manufacturers, including the Rambus licensees, are developing higher-frequency versions of standard DRAMs such as EDO, SDRAMs and SGRAMs which compete with RDRAMs. These companies are much larger and have better access to financial, certain technical and other resources than Rambus. Additional high-speed DRAMs have recently been introduced by other semiconductor companies for specialized applications.

The Company believes that its success in establishing a new high-speed memory interface has been due to the systems approach it has taken to solving the application needs of companies in home video game, PC and other electronic systems businesses. However, the Company believes competitors have begun to take a similar approach. The Company believes that its principal competition may come from its licensees and prospective licensees, many of which are evaluating and developing products based on alternative technologies. Many DRAM suppliers have indicated that they are developing a new technology called Double Data Rate ("DDR") SDRAMs, aimed at doubling the memory bandwidth from SDRAMs without increasing the clock frequency. In addition, a consortium including both large DRAM manufacturers and systems companies is promoting a specification for an alternative high-speed interface standard called SyncLink or SLDRAM. To the extent that these alternative technologies provide comparable system performance at lower cost than RDRAMs, or do not require the payment of comparable royalties, the Company's licensees and prospective licensees may adopt and promote the alternative technologies. There can be no assurance that the Company's future competition will not have a material adverse effect on the Company's business, financial

condition and results of operations. In addition, certain semiconductor companies have recently introduced a new kind of IC which combines logic and DRAM on the same chip. Such chips, called "embedded DRAM," eliminate the need for any chip-to-chip interface and are primarily being used for graphics applications. Embedded DRAMs are well suited for applications where component space saving and power consumption are important, such as in the graphics subsystems of notebook PCs. There can be no assurance that competition from embedded DRAMs will not increase in the future.

Patents and Intellectual Property Protection

The Company has an active program to protect its proprietary technology through the filing of patents. At September 30, 1997, the Company held 34 United States patents on various aspects of its technology, with expiration dates ranging from 2010 to 2014 and had applications pending for an additional 69 United States patents. The Company's United States patents do not prevent the manufacture or sale of Rambus-based ICs abroad. At September 30, 1997, the Company held six foreign patents and had additional foreign patent applications pending in Taiwan, Korea, Japan and various other jurisdictions. In addition, the Company attempts to protect its trade secrets and other proprietary information through agreements with licensees and systems companies, proprietary information agreements with employees and consultants and other security measures. The Company also relies on trademarks and trade secret laws to protect its intellectual property.

Rambus believes that it is important to develop and maintain a uniform RDRAM memory interface standard. The Company's contracts generally prevent a licensee from using licensee-developed patented improvements related to Rambus technology to block other licensees from using the improvements or requiring them to pay additional royalties related to their use of Rambus interface technology. Specifically, the contracts generally require licensees to grant to Rambus a royalty-free cross-license on patented licensee intellectual property related to the implementation of Rambus interface technology, which Rambus sublicenses to other licensees which have entered into similar arrangements. Not all licensees have granted Rambus cross-licenses and there is no assurance that such a blocking arrangement will not occur in the future.

Sales and Marketing

Consistent with the Company's business model, sales and marketing activities are focused on developing relationships with potential licensees and on participating with existing licensees in marketing, sales and technical efforts directed to systems companies. In many cases, Rambus must dedicate substantial resources to market to and support systems companies. The Company's sales and marketing efforts include applications engineering and other technical support for systems companies, as well as trade shows, advertising and other traditional marketing activities.

Employees

As of September 30, 1997 the Company had 139 employees, including three in Japan. Of this total, 99 were in engineering, 27 were in marketing and sales, and 13 were in finance and administration. Overall, approximately three quarters of the Company's employees have technical degrees, and more than half of the Company's employees have advanced technical degrees. The Company's future success will largely be dependent on its ability to attract, retain and motivate highly qualified technical and management personnel who are in great demand in the semiconductor industry. The Company's employees are not represented by any collective bargaining agreements and the Company has never experienced a work stoppage. The Company believes that its employee relations are good.

Factors Affecting Future Results

Unpredictable and Fluctuating Operating Results. Because many of the Company's revenue components fluctuate and are difficult to predict, and its expenses are largely independent of revenues in any particular period, it is difficult for the Company to accurately forecast revenues and profits or losses. Historically, contract revenues have represented the largest portion of the Company's revenues. The Company recognizes contract revenues ratably over the period during which post-contract customer support is expected to be provided. While this means that contract revenues from current licenses are relatively predictable, accurate prediction of revenues from new licenses is difficult because the development of a business relationship with a potential licensee is a lengthy process, frequently spanning a year or more, and the fiscal period in which a new license agreement will be entered into, if at all, and the financial terms of such an agreement are difficult to predict. In addition to license fees, contract revenues include fees for engineering services, which are dependent upon the varying level of assistance desired by licensees and, therefore, the revenue from these

services is also difficult to predict. Adding to the complexity of making accurate financial forecasts is the fact that certain expenses associated with a particular contract are typically front-end loaded, except for expenses associated with upgrades and enhancements, whereas contract fees associated with that contract are recognized ratably over the period during which the post-contract customer support is expected to be provided.

Royalties accounted for 22.4% of total revenues in fiscal 1997. While the Company does not expect a significant increase in royalty revenue in the near term, the Company believes that royalties will represent an increasing portion of total revenue in the long term. Increasing royalty revenues will add to the difficulty in making accurate financial forecasts. Such royalties are recognized in the quarter in which the Company receives a report from a licensee regarding the shipment of Rambus ICs in the prior quarter, and are dependent upon fluctuating sales volumes and prices of chips containing Rambus technology, all of which are beyond the Company's ability to control or assess in advance. The Company believes that its continued success will be substantially dependent upon royalties increasing at a rate which more than offsets decreases in the recognition of deferred revenue under existing contracts as their recognition periods expire, as well as the Company's ability to add new licensees and to license new generations of its technology to its existing licensees. Because a systems company can change its source of Rambus ICs at any time, and because the new Rambus license source could have a considerable nonrefundable prepaid royalty balance as well as different royalty rates, any such change by a systems company, particularly one which accounts for substantial volumes of Rambus ICs, could have a sudden and significant adverse effect on the Company's revenues.

The Company's business is subject to a variety of additional risks that could materially adversely affect quarterly and annual operating results, including market acceptance of the Company's technology; systems companies' acceptance of Rambus ICs produced by the Company's licensees; market acceptance of the products of systems companies which have adopted the Company's technology; the loss of any strategic relationships with systems companies or licensees; announcements or introductions of new technologies or products by the Company or the Company's competitors; delays or problems in the introduction or performance of enhancements or future generations of the Company's technology; fluctuations in the market price and demand for DRAMs and logic ICs into which the Company's technology has been incorporated; competitive pressures resulting in lower contract revenues or royalty rates; changes in the Company's and system companies' development schedules and levels of expenditure on research and development; personnel changes, particularly those involving engineering and technical personnel; costs associated with protecting the Company's intellectual property; changes in Company strategies; foreign exchange rate fluctuations or other changes in the international business climate; and general economic trends and other factors.

Volatility of Stock Price. The trading price of the Company's Common Stock could be subject to wide fluctuations in response to quarterly variations in operating results, announcements of technological innovations or new products by the Company, its licensees or its competitors, developments with respect to patents or proprietary rights and other events or factors. The trading price of the Company's Common Stock could also be subject to wide fluctuations in response to the publication of reports and changes in financial estimates by securities analysts, and it is possible that the Company's actual results in one or more future periods will fall short of those estimates by securities analysts. In addition, the equity markets have experienced volatility that has particularly affected the market prices of equity securities of many high technology companies and that often has been unrelated or disproportionate to the operating performance of such companies. These broad market fluctuations may adversely affect the market price of the Company's Common Stock.

History of Losses; No Assurance of Profitability. As of September 30, 1997, the Company's accumulated deficit was approximately $32.5 million. While the Company generated net income for the last four consecutive quarters, it incurred significant losses in each quarter of fiscal 1996 and in each quarter of its prior fiscal years. There can be no assurance that in the future the Company will be profitable on a quarterly or annual basis.

Dependence upon Limited Number of Licensees. The Company neither manufactures nor sells devices containing its interface technology. Rather, the Company licenses its technology to semiconductor companies, which in turn manufacture and sell Rambus ICs to systems companies which incorporate Rambus technology into their products. The Company's strategy to become an industry standard is dependent upon the Company's ability to make its technology widely available to systems companies through multiple semiconductor manufacturers, and there can be no assurance that the Company will be successful in maintaining its relationships with its current licensees or in entering into new relationships with additional licensees. The Company faces numerous risks in successfully obtaining licensees on terms consistent with the Company's business model, including, among others, the lengthy and expensive process of building a relationship with a potential licensee before there is any assurance of a license agreement with such party; persuading large semiconductor companies to work with, to rely for critical technology on, and to disclose proprietary manufacturing technology to, a smaller company such as Rambus; persuading potential licensees to bear certain

development costs associated with Rambus technology and to make the necessary investment to successfully produce Rambus ICs; and successfully transferring technical know-how to licensees. In addition, there are a relatively limited number of larger semiconductor companies to which the Company could license its interface technology in a manner consistent with its business model. The Company believes that its principal competition may come from its licensees and prospective licensees, many of which are evaluating and developing products based on alternative technologies.

Dependence upon Systems Companies. Although sales of Rambus ICs to systems companies which have adopted the Company's technology for their products are not made directly by the Company, such sales directly affect the amount of royalties received by the Company. Therefore, the Company's success is substantially dependent upon the adoption of the Company's interface technology by systems companies, particularly those which develop and market high-volume business and consumer products such as home video games and PCs. The Company is subject to many risks beyond its control that influence the success or failure of a particular systems company, including among others competition faced by the systems company in its particular industry; market acceptance of the systems company's products; the engineering, sales and marketing and management capabilities of the systems company; technical challenges unrelated to Rambus technology faced by the systems company in developing its products; and the financial and other resources of the systems company. The process of persuading systems companies to adopt the Company's technology can be lengthy and, even if adopted, there can be no assurance that the Rambus technology will be used in a product that is ultimately brought to market, achieves commercial acceptance or results in significant royalties to the Company. Rambus must dedicate substantial resources to market to and support systems companies, in addition to supporting the sales and marketing and technical efforts of its licensees in promoting Rambus technology to systems companies. To date, the Company has not charged systems companies for technical support. Because the Company does not control the business practices of its licensees, it has no ability to establish the prices at which its technology is made available to systems companies or the degree to which its licensees promote Rambus technology to systems companies.

No Assurance of Adoption of Rambus Technology as an Industry Standard. An important part of the Company's strategy to become an industry standard is to penetrate new markets by targeting leaders in those markets. This strategy is designed to encourage other participants in those markets to follow such leaders in adopting Rambus technology. Should a high profile industry participant adopt Rambus technology for one or more of its products but fail to achieve success with those products, other industry participants' perception of Rambus technology could be adversely affected. Any such event could reduce future sales of Rambus ICs. Likewise, were a market leader to adopt and achieve success with a competing technology, the Company's reputation and sales could be adversely affected. In addition, some industry participants have adopted, and others may in the future adopt, a strategy of disparaging the Rambus solution adopted by their competitors. Failure of the Company's technology to be adopted as an industry standard would have a material adverse effect on the Company's business, financial condition and results of operations.

Future Dependence upon PC Main Memory Market Segment and Intel. An important part of the Company's strategy is to penetrate the market segment for PC main memory. Rambus believes that PC main memory currently accounts for approximately one-half of all DRAMs sold. In November 1996, Rambus signed a development and license contract with Intel Corporation which provides for the parties to cooperate in the development of a specification for an extension of the RDRAM optimized for PC main memory applications. The contract also calls for Intel to use reasonable best efforts to develop a PC main memory controller designed for use with such RDRAMs. The anticipated development period for the new RDRAM technology is at least two years and there are a number of technological issues which must be successfully resolved prior to implementation. There can be no assurance that Intel will successfully develop a controller for use with RDRAMs in time to meet market requirements, or at all. In addition, there can be no assurance that the market for high bandwidth PC main memory products, as anticipated by Intel, will develop by 1999 or at all. Under the contract, Intel can terminate its relationship with Rambus at any time. Even if such development efforts are completed, there is no assurance that RDRAMs will be built by the Company's licensees and purchased by PC manufacturers in sufficient quantity to become a standard for PC main memory. The Company established a relationship with Intel several years ago, but Intel did not at that time pursue development relating to Rambus technology. There can be no assurance that Intel's current emphasis or priorities will not change in the future, resulting in less attention and fewer resources being devoted to the current Rambus relationship. Although certain aspects of the current relationship between the two companies are contractual in nature, many important aspects depend on the continued cooperation of the two companies. There can be no assurance that Rambus and Intel will be able to work together successfully over an extended period of time. In addition, there can be no assurance that Intel will not develop or adopt competing technologies in the future.

Revenue Concentration. The Company is subject to revenue concentration risks at both the licensee and the systems company levels. In fiscal 1997, 1996, and 1995, revenues from the Company's top five licensees accounted for

approximately 63%, 65% and 70% of the Company's revenues, respectively. (In fiscal 1997, NEC accounted for approximately 29% of revenues and LG Semiconductor accounted for approximately 10% of revenues.) Because the revenues derived from various licensees vary from period to period depending on the addition of new contracts, the expiration of deferred revenue schedules under existing contracts and the volumes and prices at which the licensees have recently sold Rambus ICs to systems companies, the particular licensees which account for revenue concentration have varied from period to period. These variations are expected to continue in the foreseeable future although the Company anticipates that revenue will continue to be concentrated in a limited number of licensees.

The royalties received by the Company are a function of the adoption of Rambus technology at the systems company level. Systems companies purchase semiconductors containing Rambus technology from Rambus licensees, and generally do not have a direct contractual relationship with the Company. The Company's licensees generally do not provide detail as to the identity of, or volume of Rambus ICs purchased by, particular systems companies. As a result, the Company faces difficulty in analyzing the extent to which its future revenues will be dependent upon particular systems companies.

The profitability first attained by the Company in fiscal 1997 was attributable partially to an increase in royalties from NEC, which the Company believes was largely due to royalties on Rambus technology incorporated into the Nintendo 64 video game system. For reasons described in the foregoing paragraph, the Company cannot precisely quantify this amount, because its licensees generally are not required to identify the particular products that incorporate, or the particular systems companies which purchase, Rambus ICs. The Company anticipates that sales from its licensees to Nintendo will continue to account for a substantial portion of royalties in fiscal 1998. Nintendo faces intense competitive pressure in the video game market, which is characterized by extreme volatility, frequent new product introductions and rapidly shifting consumer preferences, and there can be no assurance as to the unit volumes of Rambus ICs that will be purchased by Nintendo in the future or the level of royalty-bearing revenues that the Company's licensees will receive from Nintendo.

The Company believes its potential to generate royalties in fiscal 1998 is largely dependent on system sales by Nintendo and sales of multimedia controller chips by Cirrus Logic and Chromatic Research. None of these companies is under any obligation to continue using Rambus technology in its current product or to incorporate Rambus technology into its future products. There can be no assurance that a significant number of other systems companies will adopt the Company's technology or that the Company's dependence upon particular systems companies will decrease in the future.

Reliance upon DRAM Market; Declines in DRAM Price and Unit Volume per System. To date, a majority of the Company's royalties has been derived from the sale of logic ICs incorporating RACs. If the Company is successful in its strategy to penetrate the PC main memory market, the Company expects that royalties from the sale of RDRAMs will eventually account for the largest portion of royalties. Royalties on RDRAMs are based on the volumes and prices of RDRAMs manufactured and sold by the Company's licensees. The royalties received by the Company therefore are influenced by many of the risks faced by the DRAM market in general, including constraints on the volumes shipped during periods of shortage and reduced average selling prices. The DRAM market is intensely competitive and generally is characterized by declining average selling prices over the life of a generation of chips. Such price decreases, and the corresponding decreases in per unit royalties received by the Company, can be sudden and dramatic. Compounding the effect of price decreases is the fact that, under certain of the Company's license agreements, royalty rates decrease as a function of time or volume. With the introduction of each new generation of higher density RDRAMs, the Company generally expects higher prices resulting in higher royalties per device, but with correspondingly fewer devices required per system. There can be no assurance that decreases in DRAM prices or in the Company's royalty rates will not have a material adverse effect on the Company's business, results of operations and financial condition. There can be no assurance that the Company will be successful in maintaining or increasing its share of any market.

Rapid Technological Change; Reliance on Fundamental Technology; Importance of Timely New Product Development. The semiconductor industry is characterized by rapid technological change, with new generations of semiconductors being introduced periodically and with ongoing evolutionary improvements. Since beginning operations in 1990, the Company has derived all of its revenue from its interface technology and expects that this dependence on its fundamental technology will continue for the foreseeable future. Accordingly, broad acceptance of the Company's interface technology is critical to the Company's future success. The introduction or market acceptance of competing technology which renders the Company's interface technology less desirable or obsolete would have a rapid and material adverse effect on the Company's business, results of operations and financial condition. The announcement of new products by the Company, such as the "Direct Rambus" technology which is under development, could cause

licensees or systems companies to delay or defer entering into arrangements for the use of the Company's technology, which could have a material adverse effect on the Company's business, financial condition and results of operations.

The Company's operating results will depend to a significant extent on its ability to introduce enhancements and new generations of its interface technology which keep pace with other changes in the semiconductor industry and which achieve rapid market acceptance. The Company must continually devote significant engineering resources to addressing the ever-increasing need for memory bandwidth associated with increases in the speed of microprocessors and other controllers. Technical innovations of the type that will be required for the Company to be successful are inherently complex and require long development cycles, and there can be no assurance that the Company's development efforts will ultimately be successful. In addition, these innovations must be completed before changes in the semiconductor industry have rendered them obsolete, must be available when systems companies require these innovations, and must be sufficiently compelling to cause semiconductor manufacturers to enter into licensing arrangements with Rambus for the new technology. There can be no assurance that Rambus will be able to meet these requirements. Moreover, significant technological innovations generally require a substantial investment before their commercial viability can be determined. There can be no assurance that the Company will have the financial resources necessary to fund future development, that the Company's licensees will continue to share certain research and development costs with the Company as they have in the past, or that revenues from enhancements or new generations of the Company's technology, even if successfully developed, will exceed the costs of development.

Competition. The semiconductor industry is intensely competitive and has been characterized by price erosion, rapid technological change, short product life cycles, cyclical market patterns and increasing foreign and domestic competition. Most major DRAM manufacturers are developing higher-frequency versions of standard DRAMs such as EDO, SDRAMs and SGRAMs which compete with RDRAMs. These DRAM manufacturers, including most Rambus DRAM licensees, are much larger and have better access to financial and technical resources than Rambus. Additional high-speed DRAMs have recently been introduced by other semiconductor companies for specialized applications.

The Company believes that its principal competition may come from its licensees and prospective licensees, many of which are evaluating and developing products based on alternative technologies and are beginning to take a systems approach similar to the Company's in solving the application needs of systems companies. Many DRAM suppliers have indicated that they are developing a new technology called Double Data Rate ("DDR") SDRAMs, aimed at doubling the memory bandwidth from SDRAMs without increasing the clock frequency. In addition, a consortium including both large DRAM manufacturers and systems companies is promoting a specification for an alternative high-speed interface standard called SyncLink. To the extent that these alternative technologies provide comparable system performance at lower or similar cost than RDRAMs, or do not require the payment of comparable royalties, the Company's licensees and prospective licensees may adopt and promote the alternative technologies. There can be no assurance that the Company's future competition will not have a material adverse effect on the Company's business, results of operations and financial condition. In addition, certain semiconductor companies have recently introduced a new kind of IC which combines logic and DRAM on the same chip. Such chips, called "embedded DRAM," eliminate the need for any chip-to-chip interface and are primarily being used for graphics applications. Embedded DRAMs are well suited for applications where component space saving and power consumption are important, such as in the graphics subsystems of notebook PCs. There can be no assurance that competition from embedded DRAMs will not increase in the future.

Limited Protection of Intellectual Property; Likelihood of Potential Litigation. While the Company has an active program to protect its proprietary technology through the filing of patents, there can be no assurance that the Company's pending United States or foreign patent applications or any future United States or foreign patent applications will be approved, that any issued patents will protect the Company's intellectual property or will not be challenged by third parties, or that the patents of others will not have an adverse effect on the Company's ability to do business. Furthermore, there can be no assurance that others will not independently develop similar or competing technology or design around any patents that may be issued to the Company.

The Company attempts to protect its trade secrets and other proprietary information through agreements with licensees and systems companies, proprietary information agreements with employees and consultants and other security measures. The Company also relies on trademarks and trade secret laws to protect its intellectual property. Despite these efforts, there can be no assurance that others will not gain access to the Company's trade secrets, or that the Company can meaningfully protect its intellectual property. In addition, effective trade secret protection may be unavailable or limited in certain foreign countries. Although the Company intends to protect its rights vigorously, there can be no assurance that such measures will be successful.

Rambus believes that it is important to develop and maintain a uniform RDRAM memory interface standard. The Company's contracts generally prevent a licensee from using licensee-developed patented improvements related to Rambus technology to block other licensees from using the improvements or requiring them to pay additional royalties related to their use of Rambus interface technology. Specifically, the contracts generally require licensees to grant to Rambus a royalty-free cross-license on patented licensee intellectual property related to the implementation of Rambus interface technology, which Rambus sublicenses to other licensees that have entered into similar arrangements. Not all licensees have granted Rambus cross-licenses, and there is no assurance that such a blocking arrangement will not occur in the future.

The semiconductor industry is characterized by frequent litigation regarding patent and other intellectual property rights. While the Company has not received formal notice of any infringement of the rights of any third party, questions of infringement in the semiconductor field involve highly technical and subjective analyses. Litigation may be necessary in the future to enforce the Company's patents and other intellectual property rights, to protect the Company's trade secrets, to determine the validity and scope of the proprietary rights of others, or to defend against claims of infringement or invalidity, and there can be no assurance that the Company would prevail in any future litigation. Any such litigation, whether or not determined in the Company's favor or settled by the Company, would be costly and would divert the efforts and attention of the Company's management and technical personnel from normal business operations, which would have a material adverse effect on the Company's business, financial condition and results of operations. Adverse determinations in litigation could result in the loss of the Company's proprietary rights, subject the Company to significant liabilities, require the Company to seek licenses from third parties or prevent the Company from licensing its technology, any of which could have a material adverse effect on the Company's business, financial condition and results of operations.

In any potential dispute involving the Company's patents or other intellectual property, the Company's licensees could also become the target of litigation. While the Company generally does not indemnify its licensees, some of its license agreements require the Company to provide technical support and information to a licensee which is involved in litigation involving use of Rambus technology. The Company is bound to indemnify certain licensees under the terms of certain license agreements, and the Company may agree to indemnify others in the future. The Company's support and indemnification obligations could result in substantial expenses to the Company. In addition to the time and expense required for the Company to supply such support or indemnification to its licensees, a licensee's development, marketing and sales of Rambus ICs could be severely disrupted or shut down as a result of litigation, which in turn could have a material adverse effect on the Company's business, financial condition and results of operations.

Risks Associated with International Licenses. To date, companies based in Japan and Korea have accounted for the substantial majority of the Company's revenues, and nearly all of its international revenues. In fiscal 1997, 1996, and 1995, international revenues constituted approximately 80%, 86% and 90% of the Company's total revenues, respectively. The Company expects that revenues derived from international licensees will continue to represent a significant portion of its total revenues in the future. All of the revenues from international licensees have to date been denominated in United States dollars. However, to the extent that such licensees' sales to systems companies are not denominated in United States dollars, any royalties that the Company receives as a result of such sales could be subject to fluctuations in currency exchange rates. In addition, if the effective price of Rambus ICs sold to the Company's foreign licensees were to increase as a result of fluctuations in the exchange rate of the relevant currencies, demand for Rambus ICs could fall, which in turn would reduce the Company's royalties. The Company does not use derivative instruments to hedge foreign exchange rate risk. In addition, international operations and demand for the products of the Company's licensees are subject to a variety of risks, including tariffs, import restrictions and other trade barriers, changes in regulatory requirements, longer accounts receivable payment cycles, adverse tax consequences, export license requirements, foreign government regulation, political and economic instability and changes in diplomatic and trade relationships. In particular, the laws of certain countries in which the Company currently licenses or may in the future license its technology require significant withholding taxes on payments for intellectual property, which the Company may not be able to offset fully against its United States tax obligations. The Company is subject to the further risk of the tax authorities in those countries recharacterizing certain engineering fees as license fees, which could result in increased tax withholdings and penalties. The Company's licensees are subject to many of the risks described above with respect to systems companies which are located in different countries, particularly video game and PC manufacturers located in Asia and elsewhere. There can be no assurance that one or more of the risks associated with international licenses of the Company's technology will not have a direct or indirect material adverse effect on the Company's business, financial condition and results of operations. Moreover, the laws of certain foreign countries in which the Company's technology is or may in the future be licensed may not protect the Company's intellectual

property rights to the same extent as the laws of the United States, thus increasing the possibility of infringement of the Company's intellectual property.

Dependence on Key Personnel. The Company's success depends to a significant extent on its ability to identify, attract, motivate and retain qualified technical, sales, marketing, finance and executive personnel. Because the future success of the Company is dependent upon its ability to continue enhancing and introducing new generations of such technology, the Company is particularly dependent upon its ability to identify, attract, motivate and retain qualified engineers with the requisite educational background and industry experience. Competition for qualified engineers, particularly those with significant industry experience, is intense. The Company is also dependent upon its senior management personnel, most of whom have worked together at the Company for several years. The loss of the services of any of the senior management personnel or a significant number of the Company's engineers could be disruptive to the Company's development efforts or business relationships and could have a material adverse effect on the Company's business, financial condition and results of operations. The Company generally does not enter into employment contracts with its employees and does not maintain key person life insurance.

Management of Expanded Operations. The Company is not experienced in managing rapid growth. The Company may not be equipped to successfully manage any future periods of rapid growth or expansion, which could be expected to place a significant strain on the Company's limited managerial, financial, engineering and other resources. The Company's licensees and systems companies rely heavily on the Company's technological expertise in designing, testing and manufacturing products incorporating the Company's interface technologies. Relationships with new licensees or systems companies generally require significant engineering support. As a result, any increases in adoption of the Company's technology will increase the strain on the Company's resources, particularly the Company's engineers. Any delays or difficulties in the Company's research and development process caused by these factors or others could make it difficult for the Company to develop future generations of its interface technology and to remain competitive. In addition, the rapid rate of hiring new employees could be disruptive and adversely affect the efficiency of the Company's research and development process. The rate of the Company's future expansion, if any, in combination with the complexity of the technology involved in the Company's licensee-based business model, may demand an unusually high level of managerial effectiveness in anticipating, planning, coordinating and meeting the operational needs of the Company as well as the needs of the licensees and systems companies. Additionally, the Company may be required to reorganize its managerial structure in order to more effectively respond to the needs of customers. Given the small pool of potential licensees and target systems companies, the adverse effect on the Company resulting from a lack of effective management in any of these areas will be magnified. Inability to manage the expansion of the Company's business would have a material adverse effect on its business, financial condition and results of operations.