Rambus 3.0 – invention to products

Over 25 years Rambus has transitioned from a technology development company to a patent monetisation entity to a business focused on both licensing and product innovation. These changes have been driven from within, with the legal team playing a crucial role

While often overshadowed by a history of litigation, Rambus has always had deep roots in innovative R&D. As the company approaches its 25th anniversary, it is insightful to take a closer look at its journey from early fundamental inventions in memory technologies to asserting its patent rights and licensing and back to a focus on cutting-edge technologies and products. On the surface, this appears to be a circuitous route back to the early days of successful product launches for a company that is often known for its legal proceedings. However, take the time to examine the significant shifts in strategy and the internal and external forces behind such changes and the path back to a focus on products seems not only sound and rational, but necessary.

Rambus 1.0 – roots in products

Rambus was founded in 1990 by Mark Horowitz and Mike Farmwald, with a vision of developing innovative memory solutions to exponentially raise the performance of computer architecture. In Rambus’s early years the company collaborated with NEC, Toshiba and other customers to bring high-performance Rambus dynamic random access memory (RDRAM) memory to market, and with Nintendo and SGI to sell successful consumer entertainment products and industrial workstations.

Sony Computer Entertainment – a Rambus customer and partner – shipped over 150 million units of Sony PlayStation 2 with Rambus RDRAM memory at the computing core. During the 2001 Intel Developer Forum, Rambus customers Samsung, Toshiba and Elpida announced a ramp-up of RDRAM production to support Intel’s conversion from Pentium 3 to Pentium 4. Rambus’s high-performance memory technology provided the speed boost required for the latest generation (and fastest on the market) Intel processor architecture, operating up to 2.0 gigahertz).

As part of Rambus’s product development, the company developed many innovations in high-performance computing, including programmable read latency, variable block size, core prefetch, dual edge clocking, delay-locked loop or phase-locked loop on a DRAM, advanced power states, programmable write delay and double bus rate control, among others. These inventions were widely adopted in SDRAM, DDR, DDR2, GDDR and later standards.

Moreover, in an effort to offer, sell and further the adoption of these patented technologies, the company held Rambus developer forums from as early as 1997. These were designed to teach the industry how to incorporate these innovations into their products to ramp up overall memory and system performance.

Like any technology company making substantial investments in R&D, Rambus took appropriate measures to protect its investments by seeking patent protection for these inventions and groundbreaking innovations.

Rambus 2.0 – a changing legal and business environment

While much has been written about Rambus’s legal proceedings, this article does not focus on the company’s litigation history. There are countless articles and court filings that the reader can access for specific information regarding this aspect of the company’s past. What is relevant for this article is how this period changed the company’s business and legal strategy.

After the initial success of RDRAM in the market, and despite its superior product performance, product sales began to slump and RDRAM eventually became irrelevant as the industry moved towards the JEDEC-standardised product, SDRAM. Many reasons were posited and litigated for this change in market preference, but the reality was that RDRAM as a proprietary offering by Rambus ultimately failed in the marketplace. Notwithstanding this fact, the memory industry had been widely implementing the patented inventions in JEDEC-standardised products. With a fundamental patent position in memory, Rambus attempted to fairly license its patents to industry participants as an alternative monetisation strategy. As any IP practitioner knows, patents are distinct and separate corporate assets and patent licensing is a completely acceptable, and in some cases preferable, method of monetisation.

Ultimately, despite its good-faith and prolonged efforts to effectively license companies using its inventions, Rambus was forced to pursue litigation in the hope of receiving fair compensation for the innovations being widely practised in the industry. As every public company has a fiduciary obligation to its shareholders to maximise the value of its investments, Rambus was left with no choice than to avail of the judicial system. These legal proceedings lasted for a considerable period.

Figure 1. Rambus portfolio growth worldwide of patents and applications

160559.jpg
Innovation and technology development continues

Despite the focus on patent licensing and the fact that it was embroiled in complex and costly litigation, the Rambus innovation engine continued to develop superior memory solutions and high-speed serial link cores. These innovations were adopted and brought to market by a number of customers in their products. The innovations were also widely published and Rambus won many industry awards.

The Rambus XDR memory solution was unveiled in 2001 with groundbreaking technical features and was swiftly adopted by Sony, Toshiba, Elpida and Samsung. This formed the basis of the Sony PlayStation 3 memory architecture, a highly successful consumer entertainment system with over 80 million units shipped worldwide. Rambus has always been an active solutions provider in the serial link arena and Rambus serial link cores have been adopted by industry-leading chip manufacturers such as Toshiba, NEC, UMC, ST and EDA/IP company Cadence.

These innovations were also widely published and won many industry awards. Even while the spotlight was on litigation, the technical community continued to recognise Rambus as an innovator in memory architecture and systems.

Meanwhile, Rambus continued to file for patents related to its product R&D efforts, building a portfolio around high-performance memory and serial links based on the ingenuity of its scientists and engineers.

Despite the focus on patent licensing and the fact that it was embroiled in complex and costly litigation, the Rambus innovation engine continued to develop superior memory solutions and high-speed serial link cores

Patent value and the rise of non-practising entities

In the early to mid-2000s, there was heightened awareness of the value of intellectual property as a separate business asset disaggregated from the relevant product or business operations. Although patent licensing practices have existed since the mid-1800s, a new generation of patent licensing companies appeared during the period following the collapse of the internet bubble. During this time, in which many internet start-up companies failed, an active market developed for the liquidation of such start-ups, the patent holdings of such companies often forming their few viable remaining assets. Many start-ups had filed for internet and software-related patents, and it is this class of patent which has largely been the focus of recent discussions regarding patent licensing and litigation. With so many assets for sale, patent aggregators such as Intellectual Ventures became active buyers in the secondary market and subsequently licensed and litigated many of these patents. Others followed suit and companies were formed specifically to purchase these assets with the intention of monetising them either offensively – directly or indirectly – or defensively through subscription or membership-based services. The patent assertion entity (PAE) has now become a fixture in the legal landscape.

The market has also seen a trend for operating companies to separate their patent holdings from the operating entity. For example, Round Rock Research LLC acquired a broad portfolio of semiconductor patents from Micron during 2010 and has since licensed them to semiconductor makers. Qualcomm announced a new corporate structure in June 2012, “further protecting and insulating its valuable patent portfolio from any claims resulting from actions and activities by portions of the company other than the Qualcomm Technology Licensing Division”. Many other companies, both public and private, have separate and active patent licensing programmes.

Patent licensors

With the rise of the PAE and public outcry over litigation abuses, the White House issued a report on Patent Assertion and US Innovation (June 2013), in which it attempted to highlight the perceived problems of PAEs and the detrimental impact that they were having on the US economy. The report issued a number of executive orders and called on Congress to pass legislation to address these issues.

The report included data on the number of patent cases commenced from 2006 to 2012, a period in which patent suit filings almost doubled, pointing out in particular that “according to one estimate, 82 percent of PAE defendants were sued on the basis of a software patent”.

A recent report by the US Government Accountability Office (August 2013) highlights that “lawsuits involving software-related patents accounted for about 89 percent of the increase in defendants over this period”, (p 2) referring to the period from “2007 to 2011 [which] shows that the number of overall defendants in patent infringement lawsuits increased by about 129 percent over this period”. Many commentators have drawn a direct connection between PAE activity and software patents of questionable value granted by the US Patent and Trademark Office (USPTO) in the years preceding the collapse of the internet bubble in 2000.

Figure 2 below shows the number of patent litigation cases filed per month from 2011 to 2014. Following the passage of the America Invents Act in 2011, the number of cases filed rose from around 300 per month to around 500 per month during 2013, when much of the attention was focused on PAEs and their actions and there were widespread calls for patent reform.

However, since 2013, the rate of new filings has dropped – most notably in the run-up to and after the Supreme Court’s decision in Alice Corp v CLS Bank in June 2014, which particularly affects software-related patents.

Figure 2. Patent litigation cases filed between 2011 and 2014

160937.jpg

Recent legislative proposals

Hot on the heels of the White House paper, a number of legislative bills were proposed in Congress. Representative Goodlatte championed the Innovation Act (HR 3309), which was passed by the House of Represenatives in December 2013. This bill primarily focused on a number of key elements, including:

  • fee shifting;
  • real party in interest disclosure;
  • discovery;
  • stay of proceedings on end users; and
  • heightened pleading standards.

Of these items, fee shifting was the most controversial, but was somewhat weakened through the legislative process, as reflected in the final form approved by the House of Representatives.

Senator Leahy introduced a related bill in the Senate that focused on patent litigation. This bill also aimed to address the issue of patent reform and included many elements of the Goodlatte bill. Leahy’s bill further addressed alleged bad-faith demand letters issued by PAEs by requiring the Federal Trade Commission (FTC) to exercise enforcement authority with respect to bad-faith demand letters and treat them as unfair or deceptive acts or practices. The Leahy bill subsequently failed to gain traction in the Senate and was withdrawn in May 2014.

In February 2015 Goodlatte reintroduced a substantially identical version of his Innovation Act to the house. Given the strong support that this bill received in 2013, it is likely that patent reform will again be on the house agenda and that a related Senate version will be up for consideration later this year.

Also joining the fray, the FTC held a workshop in 2012 and announced a study in 2013 to study the impact of PAEs. Both the International Trade Commission (ITC) and the USPTO have expressed opinions on the PAE issue and adopted certain rules and procedures aimed at curtailing perceived abuses by PAEs. Finally, the federal courts have been very active in patent cases, creating new case law on attorneys’ fees, de novo reviews of claim construction, the exhaustion doctrine, indirect infringement, standard-essential patents and reasonable royalty calculations for patent damages. Any objective reading of these cases would suggest that the rulings have been unfriendly to patentees.

Given the reaction to public sentiment by Congress and the executive agencies, it has been necessary to assess the effect that this might have on a business that currently depends on patent assets and an active licensing programme. Arguably, Congress and the agencies should be focusing on software patents and players in the secondary market. However, any broad-brush policies based on rhetoric and questionable data may have a severe adverse impact on innovative companies that provide true benefits through their R&D activities. Without the licensing income, Rambus may be unable to fund these investments and provide high-paying jobs to the many scientists and engineers that it employs. It is not alone in this concern. In a June 2014 article in Forbes magazine, former USPTO Director David Kappos voiced similar sentiments regarding the protection of innovation while addressing abusive litigation tactics. To be clear, we strongly advocate for a strong patent system – including the ability to enforce valid patent rights – but also support reasonable measures to restrict abusive/bad-faith litigation, including misuse of demand letters.

Collaboration is more difficult without a value contribution, such as commercial offerings in the form of products or technology solutions

Birth of Rambus 3.0

Against this business and legal backdrop, Rambus began to undertake yet another significant strategy shift during 2012. In June 2012 Ron Black was appointed CEO and began communicating a strategy which included a strong product focus, drawing on the company’s engineering roots and well-established technical reputation. A critical element of this new strategy was not only patent licensing, but also technology and product offerings. Inherent in this strategy was the recognition that a business which depends solely on patent licensing through litigation is unsustainable.

In order to remain relevant in an industry with complex ecosystems involving multiple players – including suppliers of chips, systems on chips (SoCs), broader systems and finished consumer products – companies must collaborate. Collaboration is more difficult without a value contribution, such as commercial offerings in the form of products or technology solutions. The degree of difficulty will increase substantially where there is no exchange of value and litigation is pending between the parties. Thus, a key element of Rambus’s new strategy was to profitably resolve litigation with companies that might otherwise be partners in technology collaboration.

With this vision, and with a clear objective for regaining a relevant position within the industry, Rambus took the first step in yet another transformation through its business and legal contacts. Beginning in February 2013 and through to early 2014, the company profitably resolved all of its pending litigation by adopting a measured and fair approach to licensing its patents and technologies at market rates:

  • In February 2013 Rambus and LSI agreed a patent licence and settled all outstanding litigation.
  • In June 2013 Rambus and Hynix agreed a patent licence and settled all outstanding litigation, opening the path towards future product collaboration.
  • In June 2013 Rambus and ST settled all outstanding litigation and expanded the technology relationship to incorporate DPA countermeasures into multimedia chipsets, accelerating ST’s broad support for CryptoFirewall core technology and providing access to Rambus’s memory and serial SoC patent portfolio.
  • In December 2013 Rambus and Micron signed a broad patent cross-licence and settlement, ending many years of litigation and opening the door to future technology and product collaboration.
  • In January 2014 Rambus and Samsung – a long-time Rambus customer and technology collaborator – signed a 10-year deal and opened the door to long-term technology and product collaboration.
  • In March 2014 Rambus and Nanya reached a broad patent licence agreement and settled outstanding litigation, thus closing all existing litigation for Rambus.

Figure 3. Timeline of important Rambus events since 2013

160797.jpg

With the related licensing income from these settlements and licensing deals, Rambus secured over $1.2 billion in forward-looking revenue (approximately four times the 2013 annual revenue), providing the solid foundation necessary for the challenging transition to a more technology and product-focused company.

Figure 4 shows that the market capitalisation of Rambus has already responded favourably to these achievements.

While there are no pending actions, it would be naïve to believe that all of Rambus’s litigation is entirely behind it. While it is committed to trying to license its patents and technology to those companies using them on fair and reasonable terms, there may be instances where companies want to avoid recognising its contributions and vested property rights and simply refuse to pay. In such situations, Rambus has a duty not only to its shareholders, but also to its paying licensees – which generally include most, if not all, of the relevant industry players. As such, litigation may be necessary as a final resort.

JEDEC

In May 2014 Rambus took a decisive step to return to the memory product space by joining JEDEC Committee 40 (JC-40), which focuses on standards and performance test criteria for integrated circuits in the fast-growing cloud and server-based memory environments. This was yet another watershed moment in the transition of Rambus to a new product-centric innovation engine. This decision was the result of substantial support by both the legal and engineering groups, and the willingness of Rambus’ leadership to make a strong public commitment to its participation in standard-setting organisations for the purpose of developing new products and bringing them to market.

Figure 4. Growth in Rambus market capitalisation 2013-2015

161003.jpg
Qualcomm – a positive proof point

In June 2014 Rambus and Qualcomm signed a broad patent licence agreement covering Rambus’ memory, serial and security technologies, marking the first new customer signing following the close of all pending litigation. In addition, Rambus announced a major new product offering – the CryptoManager feature management platform – for which Qualcomm will be the lead customer. This product revolutionises the IC manufacturing process through unprecedented remote secure key provisioning and SoC configuration. This engagement with Qualcomm reflects the vision articulated by Black soon after his arrival. Patent licensing and technology offerings can and should co-exist. The Qualcomm deal is a positive proof point of this strategy.

External outreach

As part of Rambus’ focus on industry engagement, the legal group has been actively working with industry organisations on issues such as patent reform, training for patent examiners and international trade through industry bodies such as Intellectual Property Owners and the American Intellectual Property Law Association, and with the USPTO and the ITC.

Meanwhile, the innovation engine continues to hum along. Rambus scientists, engineers and researchers actively publish and present papers at industry events such as the Design Automation Conference, DesignCon and various security and imaging conferences. In addition, new technology and product offerings are being actively developed to complete the transition back to cutting-edge products which are highly relevant and widely adopted in the industry.

Rambus 3.0 and beyond

In the 25 years since its founding, Rambus has transitioned from an industry groundbreaker to a well-known litigant and back to an innovation and technology leader once again. The new mix of innovative memory, interface and security products and solutions holds the promise of a bright future of technical excellence and customer collaboration and engagement. Through it all, the legal team has been an integral part of this evolution. The company is committed to ensuring a successful business outcome for its partners, customers, employees and shareholders now and into the future.

Action plan

The Rambus business model has undergone some significant changes over the years. This experience has helped to shape the principles that guide the company today:

  • Patent licensing is an accepted and valid method of monetisation.
  • Patent licensing may also be a complement to an existing product or technology business.
  • For patent licensing and other business lines to co-exist, patent licensing practices must take into account the need for technical collaboration.
  • Understanding market rates will likely further efforts to close patent licensing deals necessary for further technical collaboration.
  • Despite good-faith and best efforts, litigation may be necessary to enforce patent rights.
  • IP professionals must understand the current legal, political and business context and environment, as well as any changes that may affect patent rights.
  • The legal department and IP professionals are critical to a company’s corporate strategy and must be intimately involved in planning and execution to ensure success of the organisation.

Michael T Moore is vice president, intellectual property and deputy general counsel, and Jae Kim is senior vice president and general counsel, Rambus, Sunnyvale, California

Unlock unlimited access to all IAM content