Intel takes chipset dispute with Nvidia to court
Updated at 9:15 a.m. PST with official comment from Intel.
Bit-tech.net tipped us off Wednesday morning that Intel has sued Nvidia over the latter's right to create and sell motherboard chipsets that support Intel's Nehalem (aka Core i7) class of desktop processors.
We haven't heard much about this fight since May 2008, although since that time Nehalem has come to market and Intel remains the only manufacturer with a supporting chipset.
We spoke to Derek Perez, Nvidia's head of public relations, Wednesday morning, and he provided some clarity on the nature of the court filing, as well as Nvidia's take on the filing, which he says is actually a request for an injunction to prevent Nvidia from manufacturing a Nehalem chipset.
"We have a cross-licensing agreement with Intel, entered into about four and a half years ago. Intel is now basically saying the cross-license agreement doesn't apply to future bus interfaces, specifically DMI, (the direct media interface Intel uses to link the Nehalem CPU to a system's memory, a new feature for Nehalem chipsets). Intel has now filed an injunction against us, basically trying to stop us from innovating on DMI.
"We believe our business license with Intel lets us build chipsets for Intel CPUs," Perez said, "specifically for CPUs with integrated memory controllers (such as DMI). The PC has become a GPU-based platform as much as a CPU-based platform, and Intel is trying to delay that inevitable shift. They're trying to do everything they can to slow that down and it's because they recognize the importance of the GPU."
This sentiment was echoed by Nvidia's official press release on the court filing, which hit our in-box while we had Perez on the phone.
"'We are confident that our license, as negotiated, applies," said Jen-Hsun Huang, president and CEO of Nvidia. "At the heart of this issue is that the CPU has run its course and the soul of the PC is shifting quickly to the GPU. This is clearly an attempt to stifle innovation to protect a decaying CPU business."
Among other things, Intel makes CPUs and the circuitry (aka "chipsets") that goes in the motherboards to which those CPUs connect. Nvidia makes 3D cards (aka "GPUs," for graphics processing unit), and it has licensed chipset technology from both AMD and Nvidia to make its own chipsets that support each CPU vendors' respective processors.
Within the last year or so, Nvidia has also aggressively pitched its GPUs as more than simply gaming hardware. By virtue of its CUDA software, and others' software that performs a similar task, Nvidia has begun encouraging software developers to offload certain programming tasks (generally multimedia-oriented operations) from the CPU to the GPU. If this type of programming becomes widespread, Nvidia can then argue that the graphics processor is more important than the traditional CPU to the way we actually use our computers.
Whether Intel's CPU business is decaying is up for debate. Financially speaking, recent shipment data from analyst house IDC indicates that while Intel is maintaining its CPU market share, its revenues are down from 2007, and the outlook for the first half of 2009 is similarly disheartening.
Despite that bad news, a recent announcement that Intel is ramping up its chip production schedule indicates that Intel is willing to take aggressive steps to further its market lead over AMD. This production ramp will also speed the introduction of new mobile chips that include a built-in GPU, a competitive move against Nvidia as well.
For Nvidia's part, its business has also been challenged lately, with layoffs in September 2008, pay-cuts, and a similarly tough outlook for the coming year. And while Nvidia's charge that Intel's CPU business is "decaying" is largely gamesmanship, if financial performance is the primary indicator of healthy technology, Intel could just as easily make the same accusation about Nvidia.
Here's Intel's official comment: "Intel has filed suit against Nvidia seeking a declaratory judgment over rights associated with two agreements between the companies. The suit seeks to have the court declare that Nvidia is not licensed to produce chipsets that are compatible with any Intel processor that has integrated memory controller functionality, such as Intel's Nehalem microprocessors and that Nvidia has breached the agreement with Intel by falsely claiming that it is licensed.
"Intel has been in discussions with Nvidia for more than a year attempting to resolve the matter but unfortunately we were unsuccessful. As a result Intel is asking the court to resolve this dispute. It is our hope that this dispute will not impact other areas of our companies' working relationship."
Rich Brown reviews desktops and various other components and peripherals for CNET. E-mail Rich. 
From that perspective, I still find Nvidia's statement presumptuous (and above all, as I said previously, gamesmanship). The BadaBoom! video app (which Nvidia more or less commissioned), a few Photoshop functions, and a handful of other programs can all ping the GPU. And when DirectX 11 brings its native GPU compute code, we may see even more apps that regularly depend on the GPU for parallel processing. That said, I don't believe we've reached the point where the GPU has surpassed the CPU's overall contribution to a computer's ability to run popular software.
I say this because most people aren't hardcore PC gamers, nor do they edit HD video or even really dig into Photoshop. When the core functionality of things like Facebook, lightweight photo editing apps, general Web browsers, and media players benefit from a GPU, then Nvidia might have an argument.
I will like to read the legal papers of both parties.
the first would be that the EU fines Intel a few billion euro's for denying competition
The second being NVidia saying 'well you can't use SLI in your Chipsets' and then AMD/ATI telling Intel that they can't use CrossfireX in their Chipsets which will end in Consumers going for AMD processors and AMD socket motherboards for SLI and CrossfireX.
The third would be that NVidia stops allowing them SLI in their chipsets and ATI makes profit because New Gaming PC builds will go for an Intel/CrossfireX solution.
But I don't think that CPUs are going to be obsolete any time soon, nor for the years to come, so speculating about Intel's financial situation is pointless. The GPU and CPU have very distinct functions, which probably won't change soon.
[quote]From that perspective, I still find Nvidia's statement presumptuous (and above all, as I said previously, gamesmanship). The BadaBoom! video app (which Nvidia more or less commissioned), a few Photoshop functions, and a handful of other programs can all ping the GPU. And when DirectX 11 brings its native GPU compute code, we may see even more apps that regularly depend on the GPU for parallel processing. That said, I don't believe we've reached the point where the GPU has surpassed the CPU's overall contribution to a computer's ability to run popular software.
I say this because most people aren't hardcore PC gamers, nor do they edit HD video or even really dig into Photoshop. When the core functionality of things like Facebook, lightweight photo editing apps, general Web browsers, and media players benefit from a GPU, then Nvidia might have an argument.[/quote]
I'm a NVIDIA partner and based on the training information I've seen, nothing suggests that NVIDIA is marketing the GPU as a replacement for the CPU. Instead NVIDIA's approach is to use the GPU's compute advantage (1400 compute cores vs. the 1 - 4 compute cores in standard CPUs) for applications that can be benefited by it.
They don't argue that applications would have to be written specifically to take advantage of it - that's the point of the whole CUDA campaign and their partnerships with Adobe and Folding@Home. Applications written to tap into the compute power of NVIDIA GPUs can process calculations much faster than a CPU that is multi-tasking. Using NVIDIA SLI technology, potentially applications can tap into 100x or more compute power than with CPUs alone, hence the performance numbers seen when using CUDA optimized applications. Another big advantage is the fact that CUDA is Open Source. Anyone can write an application that takes advantage of using the CUDA technology by coding using C and the CUDA API or OpenCL.
There's a reason Apple is transitioning to NVIDIA based chipsets over Intel only based chipsets, and the additional compute power is a significant reason why.
But as quoted in the original post, Nvidia's CEO Jen Hsun Huang also said in Nvidia's official, widely distributed press release: "At the heart of this issue is that the CPU has run its course and the soul of the PC is shifting quickly to the GPU."
I think most people would agree that GPUs and CPUs each have their strengths, and the role of the GPU as a generalized parallel processor will likely increase in the near future. But let's not pretend there isn't a marketing game going on here as well.
what era are you stuck in?
ATi is very competitive now, they don't have the performance crown now, but the 4870X2 offers performance near the GTX295 for significantly less, HD4890 is moving to 40nm, nVidia just moved to 55nm.
Intel makes the graphics products they do for a reason. Lower cost, cheaper to produce, uses less battery power, or whatever. That's fine and I'm perfectly happy with that. If I want something more powerful I'll buy an NV card. I'm not complaining about it. But it sounds like Intel is. If they're not happy with the products they offer then offer different products. Suing NV out of the game isn't good for anyone. Good old competition is.
The EU has put a lot of pressure on Microsoft, to open up Windows' interfaces, so competitive software products can interwork properly. This certainly sounds like a comparable anti-competitive situation between Intel and Nvidia.
Sorry but I don't see how what they are doing is a problem. Tell me which motherboard now supports DDR4 (assuming they continue in order here) memory. Then tell me which board can support a P4 and DDR3 memory. Don't mean to come off with an attack but that's just kind of the price people and manufacturers have to pay to get new technology. If people are die hard technology fanatics i'm sure they wont mind paying to upgrade memory/cpu/and mobo when the newer (and overpriced might I add) memory technology comes out. Ya there may be a delay when DDR4 is offered to when Intel releases a compatible CPU but i'm sure they'll be quick to make one. But again this is just the price everyone pays for so many technology markets being employed on a single computer. All of your other statements seem valid though :-P