Q: Give us an overview of Rambus' IP portfolio. Does Rambus plan to diversify its portfolio beyond memory and semiconductor? If so, what new areas will be addressed?

A: The heart of our portfolio, built up over our two-decade history, is patented inventions for memory and serial links. These enable both industry-standard solutions such as the double data rate (DDR) and low-power DDR (LPDDR) family of memories and serial links, including Peripheral Component Interconnect (PCI) Express as well as our leadership solutions such as XDR and Mobile XDR memory and the FlexIO interface.

In support of our business, we've created a world-class innovation, patent development and licensing platform that can extend to other technology areas. Building on this platform, we have an active diversification strategy. While we continue to invest in memory architectures and serial links, we're expanding into new areas in the semiconductor space such as imaging, as well as into technology areas beyond semiconductors, including light-emitting diode (LED)-based lighting and displays, and electronic system security with our acquisition of Cryptography Research.

Q: With chip design costs continuing to escalate at an alarming rate, the need for first-time-right silicon has never been more imperative. How does Rambus' IP enable its customers to meet this significant milestone?

A: While the need for first-time-right silicon is an economic imperative, achieving that goal is increasingly difficult. A major driver of design costs is the rapidly rising design complexity with each new process node. To ensure the robustness of our solutions, we use a total system design approach. At Rambus, we've brought together experts in system and device architecture, analog and mixed-signal design, signal and power integrity, as well as board and package design to develop our solutions.

We validate the performance of our solutions through silicon test vehicles. As an example of what we offer to our licensees, our physical layer (PHY) development packages include all the necessary building blocks, PHY architecture, schematics, models, generic layout, floor plan, verification IP, implementation documentation, testing documentation, design scripts and simulation files to ensure design success.

Q: Over the last decade, system-on-chip (SoC) devices have required enormous amounts of internally and externally developed semiconductor IP cores. Never before have semiconductor companies relied so heavily on their suppliers. Not only is the partnership between semiconductor companies and their semiconductor IP suppliers an important one, it is also a complex one. What are some of the complexities that exist within this partnership?

A: The increasing reliance on semiconductor IP providers is another reflection of rising design complexity. No one company has the resources to be best-in-class in every technology necessary for today's highly integrated SoC devices. Semiconductor IP partners can provide the domain expertise to deliver SoC companies with leading-edge technology, allowing the SoC companies to focus on their core competencies and competitive differentiation in chip-level solutions.

One of the greatest challenges for SoC companies is the integration of IP from many partners. Using a total system design approach, we ensure our solutions are implemented with significant margin for successful integration. We also put a great deal of emphasis on developing the tools and materials to support integration. While each engagement is unique, typically our engineers work hand-in-hand with those of our licensees to ensure successful integration of our solutions into their semiconductor products.

I think an even greater success factor for our licensees is that we focus on designing our solutions to meet the needs of the end system. SoC companies that work with us have a partner who has validated the desired features and requirements with system customers and designed our solution accordingly. In this way, we help ensure the success of our licensees' SoCs in end products. This is one of the most frequent compliments we hear from our customers.

Q: Tell us about some of Rambus' most significant patent license agreement partnerships.

A: Toshiba and Panasonic have been licensees since the early days of Rambus. Toshiba, in fact, was our very first licensee. Our technology has enabled tens of millions of Toshiba and Panasonic HDTVs, PCs, and DVD and Blu-ray players.

Sony's PlayStation^®2 and PlayStation^®3 are flagship applications for Rambus technology. The PS2 has sold over 150 million units worldwide and the PS3 over 50 million units. In support of these two platforms, we've licensed Rambus solutions to Toshiba, IBM, Elpida, Samsung, TI, IDT and Cypress.

In 2010, we signed new agreements with both Samsung and Elpida, which were major milestones for Rambus. These agreements are further validation of the value of our innovations. We have great opportunities to work with two renowned leaders in the memory market on exciting new solutions for next-generation systems.

Q: Participants of GSA's IP Licensing Survey were invited to GSA's Spring IP Summit to discuss what is working and not working in IP licensing. Items such as "acceptance of the IP model" and "low barriers to entry for new companies" were mentioned as "working" and "customers' willingness to pay royalties" and "suppliers' ability to accurately convey the value of their IP" were mentioned as "not working." What other views does Rambus have on what is working and not working with IP licensing?

A: We're obviously strong believers in the IP business model. We think it's a great way to foster innovation and promote the broad adoption of best-in-class technology across a range of products.

However, it's not for everybody. When you're developing IP for next-generation applications, or even further out in the future, it takes a long time before that IP is ultimately shipping in products and can be monetized through royalties. It takes a strong financial position to fund innovation efforts that create leading-edge IP and to fund the engineering resources needed to support licensees.

Q: Many in the industry hold the view that memory chips are a commodity product. Does Rambus hold this same view? If not, how do Rambus' memory architectures fight the commodity status?

A: Most high-volume memory chips, like other semiconductor products, move from differentiated to commodity status during their product life cycle. We believe there are some great opportunities for differentiation in memory products in applications ranging from smartphones and tablets to servers. We focus on enabling our licensees to create differentiated products with our innovative features and leadership architectures such as Mobile XDR memory.

In the years ahead, traditional memory technology is going to run into increasingly severe scaling limits. The introduction of new materials and architectures will offer further opportunities for differentiation.

We've got an ongoing commitment to innovation that advances the frontier of memory performance. Employing our patented innovations or our leadership solutions built on these innovations, our licensees can create differentiated memory products that deliver great value for electronic system companies and consumers.

Q: Looking ahead, do you see 3D stacking technology as part of the future of memory technology? What concerns does Rambus have about the revolutionary technology?

A: The industry is undergoing a revolutionary change in system-level requirements, especially for mobile applications, while the prevailing preference is to stick to evolutionary solutions which conserve infrastructure, manufacturing methods and supply chains. One of the key benefits of 3D stacking in memory technology is to improve power efficiency while increasing bandwidth, but it comes with the need for significant changes to business models and supply chains. We believe there are ways to apply new signaling technologies and architectural enhancements to existing infrastructure to greatly improve memory and system performance and power efficiency.

At Rambus, we have been studying 3D stacking techniques for a number of years and have developed innovations which we believe will ultimately enable access to the benefits of the technology while minimizing the downsides. There's great value for 3D stacking down the road. To realize that value, there's a lot of work that we as an industry must do to get it ready for high-volume production in the second half of this decade.

Q: As of June 30, Rambus had acquired two companies in 2011—Imagine Designs and Cryptography Research. What was Rambus' expectation of these acquisitions? Does Rambus plan to make M&A activity a significant part of its five-year strategy?

A: Rambus is a company of inventors. It is a great place for engineers and scientists to work on the frontiers of technology. We are continually looking for opportunities to bring on board key inventors and key technologies that support our mission of continuously enriching the end-user experience of electronic systems. We believe this is critical to our continued success.

The acquisition of Imagine Designs added some great lighting technologies to our Lighting and Display Technology business. In addition, Brian Richardson, Imagine Design's former president and principal inventor, joined Rambus to continue developing his inventions and solutions.

Our purchase of Cryptography Research, which is now a division of Rambus, added a completely new set of technologies to the Rambus portfolio. Cryptography Research's president and principal inventor Paul Kocher and his entire team of engineers and scientists have joined Rambus to continue innovating in the important area of semiconductor and electronic system security.

Our expectations of our Lighting and Display Technology business, of which the Imagine Designs technology is now a part, and our Cryptography Research business, is that these businesses will contribute significantly to Rambus' revenue in the years ahead.

Back to Industry Reflections Home