Intel, NVIDIA Swap Patents, But Who Gains What?
"This agreement ends the legal dispute between the companies, preserves patent peace and provides protections that allow for continued freedom in product design," said Doug Melamed, Intel senior vice president and general counsel. "It also enables the companies to focus their efforts on innovation and the development of new, innovative products."
Nvidia is similarly pleased: "This agreement signals a new era for NVIDIA," said Jen-Hsun Huang, NVIDIA's president and chief executive officer. "Our cross license with Intel reflects the substantial value of our visual and parallel computing technologies. It also underscores the importance of our inventions to the future of personal computing, as well as the expanding markets for mobile and cloud computing."
It's not precisely clear what the two companies have agreed to cross-license. We're assuming that NV has granted Intel a license to use SLI in whatever chipsets it wants; Jen-Hsun's comments further imply that Intel gained access to some of Nv's mobile technology. Given how many mobile projects NV is currently working, that could be be a reference to Tegra, Denver, or even Ion (though that's doubtful).
We do know something about what NV got in return. Those of you hoping this means NV has at last gained its own x86 license are in for a disappointment, Intel's first comment is: "NVIDIA receives a license to Intel's patents subject to the terms of the agreement, including that x86 and certain other products are not licensed to NVIDIA under the agreement." (emphasis added).
Nvidia's press release tells us a bit more. "NVIDIA will...retain use of Intel's patents, consistent with its existing six-year agreement with Intel. This excludes Intel's proprietary processors, flash memory and certain chipsets for the Intel platform." (Emphasis added). That doesn't leave much room where Intel's core competencies are concerned, but it could include power-saving technologies, SoC integration-related patents, or access to other Intel microcontrollers like south bridges, NICs, or WiFi adapters.
NV also held a brief webcast this afternoon to discuss the deal, but neither the comments nor the Q&A revealed much of additional interest. During the session, Jen-Hsun noted that the arrangement allows Intel to integrate certain unspecified GPU technologies into a CPU like Sandy Bridge, but didn't go into additional detail.
An Intel/Nvidia collaboration could give Santa Clara a huge leg up when it comes to competing with AMD's integrated GPUs, but we're not willing to jump to that conclusion just yet. Not only are there a wide range of potential platforms involved, it'll be quite some time before we see the fruits of this collaboration in any form. Even if Intel wants to integrate an NV GPU into future Core products, it'll need to adapt NV GPUs to its unique integration and shared cache approach. Intel might be able to sneak the new technology into Ivy Bridge (though that seems doubtful), but we're probably talking about Ivy Bridge's successor, Haswell and 2013 at the earliest.
As for what NV got out of the deal, company executives were unwilling to comment on that situation past what we've already described. Hopefully that's not because NV got a weak hand (or a weak per-GPU price in a scenario like the above). $1.5 billion is a tidy amount of cash, but it won't amount to much more than a bit of padding compared to the company's multi-billion dollar revenue throughout the year.