Nvidia Firmly Denies Plans To Build a CPU

Barence writes "A senior vice president of Nvidia has denied rumours that the company is planning an entry into the x86 CPU market. Speaking to PC Pro, Chris Malachowsky, another co-founder and senior vice president, was unequivocal. 'That's not our business,' he insisted. 'It's not our business to build a CPU. We're a visual computing company, and I think the reason we've survived the other 35 companies who were making graphics at the start is that we've stayed focused.' He also pointed out that such a move would expose the company to fierce competition. 'Are we likely to build a CPU and take out Intel?' he asked. 'I don't think so, given their thirty-year head start and billions and billions of dollars invested in it. I think staying focused is our best strategy.' He was also dismissive of the threat from Intel's Larrabee architecture, following Nvidia's chief architect calling it a 'GPU from 2006' at the weekend."

Comments

Inaccurate headline

[TheRaven64]

nVidia are building a CPU, a Cortex A9 derivative with a GPU on-die and a load of other nice features. The summary states that they're not building an x86 CPU, but this is not what the headline says.

Anyone Surprised?

[Underfoot]

Is anyone actually surprised that the CEO is denying this? Even if the rumors were true, letting news out to market about it would give Intel time to prepare a response (and legal action).

Re:Anyone Surprised?

[morgan_greywolf]

I don't get the legal action part. Is the x86 architecture patented by Intel? Even if it is, wouldn't the patent have expired by now? After all, its more than 30 years old. Do AMD, VIA etc. pay licensing fees to Intel for building processors using the x86 architecture? If so, why cant NVidia?

Yes. Various pieces of parts of the x86 architecture that have been developed within the last 20 years (noteably, stuff related to the IA32 architecture of the 386, 486 and Pentium and later lines) are all still under patent.

Patents filed before June 8, 1995 get the greater of 17 past the patent grant date or 20 years total, whichever is greater.

Re:Anyone Surprised?

[AKAImBatman]

Is anyone actually surprised that the CEO is denying this?

Not at all. As you say, he would have denied it even if NVidia WAS planning a CPU. What actually speaks volumes IMHO, is the vehemence with which he denied it. Any CEO who's cover-denying a market move is not going to close his own doors by stating that the company could never make it in that space. He would give far weaker reasons so that when the announcement comes the market will still react favorably to their new product.

In other words: stick a fork in it, because this bit of tabloid reporting is dead.

Re:Anyone Surprised?

[Daengbo]

Check the URL before clicking.

Reprogrammable GPU?

[Wills]

When hell freezes over, they could release a GPU where the instruction set is itself microprogrammable with open-source design, and then end users could decide whether they want to load the GPU's microcode with an x86 instruction set, a dsp set, or whatever.

Re:Reprogrammable GPU?

[HerculesMO]

If hell froze over they wouldn't have to worry about the cooling on their chips.

I guess that's a plus.

Re:Reprogrammable GPU?

[Fizzl]

And I want a microwave than can be customer bludgeoned into a bicycle. Where do you people get the idea that you can do hardware in software?

Re:Reprogrammable GPU?

[Kjella]

Who said price is the most interesting issue? I'd definitely choose the versatility of an open-source microcode GPU that could be dynamically reprogrammed to have any of several different instruction sets.

As long as they're Turing complete, any of them can in principle do anything. Yes, then at least to me it comes down to price - if it's cheaper to have a car, boat and plane than making a tranasformer that can do all three at it, suck at all three and cost a bajillion more I'll go for traditional chips, thank you very much.

Re:Reprogrammable GPU?

[MarcQuadra]

Transmeta tried that. It was slow, expensive, and inconsistent. Also, nobody ever used any other 'instruction sets' besides x86, mostly because that's the most-common-denominator in the computing world.

It sucks, it's not the -best- way to do it, but it's the way the market seems to favor. Just ask Apple, Sun, DEC, and HP.

Focused

[Akita24]

Yeah, they've stayed focused on graphics chips, that's why there are so many motherboards with nVidia chip sets .. *sigh*

Re:Focused

[microbrew_nj]

I can think of a few good reasons for Nvidia to roll their own chipsets. SLI is one. The market for integrated motherboards (with their chipset) is another.

Just a thought...

[darkvizier]

If you're 30 years behind them in their market, and they're 2 years behind you in yours, maybe it's not wise to be "dismissive of the threat" ?

rumour machine

[Anonymous Coward]

rather handy that this rumour gives nvidia, a GPU company, the chance to point out how futile it would be for them to try and enter the CPU market... then point over to intel, a CPU company, trying to make a GPU...

Difficult

[DrYak]

Microcode-upgrade are possible for CPU that have a huge big complex reprogrammable pipeline like the current top of the line CPUs, or CPU where the pipeline is handled in software (like the Transmeta chips).

GPU, on the other hand, have a very short and simplistic pipeline which is hard-fixed. They draw their tremendous performance, from the fact that this pipeline drives ultra-wide SIMD units which process a fuck-load of identical threads in parallel.

But there nothing much you could reprogramm currently. Most of the die is just huge cache, huge registry files, and a crazy amount of parallel floating point ADD/MUL blocks for the SIMD. The pipeline is completely lost amid the rest.
(Whereas on CPU, even if the cache dwarfs the other structure, there are quite complex logic blocks dedicated to instruction fetching and decoding).

Re:Difficult

[billcopc]

Let me guess: you've never read anything about microprocessor engineering, have you ?

What you describe is what every non-engineer dreams of. You want a chip that any idiot can reprogram, without knowing the "less simple" ways of FPGAs. That's kind of like saying you want a car that gets 200 miles to the gallon, can park in a shoebox and carry 20 kids in the back seat - oh, and it drives itself automagically so your kids can take themselves to soccer practice without bugging you.

The reason why no one ever builds such monstrosities is because there is simply no point to it, when you can have purpose-built chips designed and fabbed for a fraction of the cost. People don't stop breathing just because their device needs 2 distinct chips instead of one jesus-truck.

x86 rumors origin ?

[DrYak]

Currently nVidia is partnering with VIA for small form factor x86 boxes. And they have made several presentation about a combination of (VIA's) x86-64 Issaiah and (their own) embed GeForce.
Touting that the platform would be the first small form factor able to sustain Vista in all DX10 and full Aero glory.

Maybe that is where some journalist got mixed and where all this "nVidia is preparing a x86 chip" rumor began ?

Re:x86 rumors origin ?

[CodeBuster]

Maybe that is where some journalist got mixed and where all this "nVidia is preparing a x86 chip" rumor began?

This is what happens when technical information is filtered through the brain of a salesperson, manager, or executive. It comes out completely mangled on the opposite side or, even worse, it morphs into something which while technically correct is NOT the information that the non-technical person thought they were conveying (i.e. they have unknowingly modified the requirements specification in a way that is logically consistent from a technical standpoint, but will result in the wrong product being built).

And why not?

[geogob]

I wouldn't mind seeing more players in the computer processor industry. The headlines really make it sound like it would be a bad thing. Maybe I'm getting the headlines wrong, but having Nvidia presenting new alternatives to a market almost exclusively owned by Intel and AMD would be interesting.

From 2006

[Alioth]

"A GPU from 2006" sounds a lot like famous last words.

I wonder if anyone at DEC made comments in a similar vein about Intel CPUs, when the Alpha was so far ahead of anything Intel was making? NVidia's architect should not underestimate Intel, if he does, he does it at his company's peril.

Re:From 2006

[Lumpy]

The alpha failed because the motherboards were $1300.00 and the processors were $2600.00 nobody in their right mind bought the stuff when you could get Intel motherboards for $400 and processors for $800.00 (dual proc boards, high end processors)

DEC died because they could not scale up to what the intel side was doing. you had thousands of motherboards made per hour for Intel with maybe 4 a day for Alpha. It's game over at that point.

I loved the Alphas, I had a dual alpha motherboard running windows NT it rocked as a server.

Re:wouldn't this be a good thing?

[ThisNukes4u]

Actually nvidia doesn't own any fabs, they contract out all their chips to TSMC, same as ati. Although now ati/amd are going to be making their fusion chips at TSMC, so they will definitely have the expertise to make x86 chips in the near future(TSMC will).

How nVidia "Survived"

[Bruce Perens]

I think the reason we've survived the other 35 companies who were making graphics at the start is that we've stayed focused.

3DFx was the first company to publish Open Source 3D drivers for their 3D cards. nVidia sued them, then bought them at a discount, and shut down the operation. So, we had no Open Source 3D for another 5 years.

That's not "staying focused". It's being a predator.

Bruce

Re:How nVidia "Survived"

[Rufus211]

What on earth are you talking about? 3DFx died because it was horribly mismanaged and ran out of money. There were lawsuits, but 3dfx sued NV first in 1998 and then in 2000 NV counter-sued (source). True NV's countersuit was right before 3dfx died, but a simple lawsuit that's gone nowhere in the courts yet doesn't cause a company to go bankrupt overnight.

Personally I'll believe one of my (ex-3dfx Austin) friend's explanation for their downfall: the fully stocked Tequila bar that was free to all employees. Or there's a whole list of problems leading to their decline on wikipedia.

Re:How nVidia "Survived"

[alen]

3dfx's problem was they could never figure out how they sold their cards. they flipped flopped from themselves to having others make the cards like Nvidia does. after so many times no one wants anything to do with you because it's bad for business planning.

nvidia has had it's current selling model for 10 years and only its partners have changed. if you want to sell video cards you can trust that if you sell cards based on nvidia's chips they won't pull the rug out from under you next year and decide to sell the cards themselves

Re:How nVidia "Survived"

[Bruce Perens]

Pixar had an OEM model too, back in its days of making hardware and software products (the Pixar image computer, Renderman, Renderman hardware acceleration) while waiting for the noncompete with Lucasfilm to run out. It's a very difficult way to run a business, because you have to pull your own market along with you, and you can't control them.

It does look like 3DFx bought the wrong card vendor. They also spun off Quantum3D, then a card vendor, which is still operating in the simulation business.

Re:Only reason

[Lisandro]

How is a "GPU" different from a "CPU"?

The GPU is a specialized (vector) processor, while the CPU is a general purpose one. What the GPU does, it does great. But its reach ends pretty much there.

The nVidia is programmed with a specific higher-order assembly language, We rely solely on the hardware vendor for tools. I think that this is UNIQUE in the (mass-market) processor world. And this is why Intel, with an x86 compatible GPU is such a threat.

You're confused. Intel is not working on a "x86 GPU". Intel is working on a new GPU design - the kicker being that this is a relatively high performance one, instead of the kind of GPUs they offered so far (feature packed, but lacking in performance). The x86 instruction set has nothing to do with it, and in fact, has nothing to do with GPU programming, which is a completely different beast.

Can anyone else produce an OpenGL shader compiler for the nVidia? Or, better yet, extend it to do NON-shader tasks. How about for the AMD?

If i'm no mistaken, nVidias CG compiler is now open sourced. So yes.