Slashdot Log In
Nvidia Rumored To Be Readying X86 Chip Release
Posted by
timothy
on Wed Aug 20, 2008 02:31 PM
from the probably-not-for-wristwatch-computers dept.
from the probably-not-for-wristwatch-computers dept.
jdb2 writes with the (honestly labeled) rumor from the Inquirer "that Nvidia is preparing to release an x86 microprocessor with its guns targeted directly at its two major rivals — Intel and AMD/ATI," and excerpts from the just-linked Inquirer article: "THE HOT RUMOR going around IDF ... [is] that the company will do an x86 part. The background whispers say that the part will be announced next week at Nvision ... Nvidia's men in white coats certainly have the brainpower to do it, but they also most certainly don't have a license to sell such a part. NV is basically locked out unless Intel and AMD both decide to be magnanimous, and we would not recommend holding your breath waiting for this to happen ... That leaves the lawsuit option open ... Any attempt to enter the market without a license would bring down Intel legal on them like flying monkeys blackening the sky. It would get ugly. Really ugly. Expensive too.""
Related Stories
This discussion has been archived.
No new comments can be posted.
The Fine Print: The following comments are owned by whoever posted them. We are not responsible for them in any way.
Full
Abbreviated
Hidden
Loading... please wait.
Odd (Score:4, Insightful)
Didn't i just read that nVidia was getting out of the x86 chipset business? Why would they now be releasing an actual x86 Chip if they don't want to even be in the chipset business?
http://tech.slashdot.org/article.pl?sid=08/08/02/1749213 [slashdot.org]
Re:Odd (Score:5, Informative)
Parent
Re: (Score:3, Interesting)
Nvidia also denied that rumour vigorously, going so far as to demand a retraction of the story (from the news site, not Slashdot...). As typical, everyone seems to have caught the rumour and completely missed the denial.
Right. Because when companies issue a denial of a rumour, they're always telling the truth. They'd have no reason to cover up some new product they're not quite ready to release ...
Oh, never mind, this is Nvidia, not Apple. Carry on.
Re:Odd (Score:4, Insightful)
So based on past performance, I would say that the Nvidia denial is correct, and the rumour is false - we aren't talking about a denial to cover up a new product, we are talking about a denial of a rumour that could cost Nvidia significant stability and market confidence. In my opinion, whomever spread the rumour should be investigated by the SEC or whoever else has jurisdiction.
But anyhow - why should we put more weight on the rumour and dismiss the denial as you seem wanting to do?
Parent
Re:Odd (Score:5)
Parent
Re:Odd (Score:5, Informative)
Yes you did, and if you'd kept reading you'd have seen that story debunked - it's Via that's getting out of the chipset business, not Nvidia.
Parent
Re: (Score:3, Funny)
You did just read that, however that was also a false rumor.
Re:Odd (Score:4, Informative)
Perhaps its the 3rd party chip-set business they don't want to be in.
Parent
Okay, I'll bite... (Score:5, Interesting)
How is it that AMD is able to release x86 chips, but nVidia can't without a license from Intel? Why would nVidia need AMD to be gracious?
Re:Okay, I'll bite... (Score:5, Interesting)
In the olden days, chip consumers insisted on a second source. AMD was annointed as Intel's second source so that Intel could sell to such folks (like the US government of yesteryear).
Parent
Re:Okay, I'll bite... (Score:5, Informative)
In the olden days, chip consumers insisted on a second source. AMD was annointed as Intel's second source so that Intel could sell to such folks (like the US government of yesteryear).
That's how AMD got the schematics to the original 8086, but that's no longer very relevant. Much more important today is AMD's patent cross license agreements with Intel. (BTW, the cross licensing also helped save Intel's position in the marketplace because it entitled them to use AMD's X86-64 design verbatim after the Itanium fiasco.)
Parent
Re: (Score:3, Interesting)
"Much more important today is AMD's patent cross license agreements with Intel."
Indeed, I did not mean to suggest that things have remained as they were in the 8086 days; just provided the origin. There is a long and tangled history of licensing between AMD and Intel. No doubt the best bits aren't public info anyway (although I suppose combing through the various legal filings in various suits could prove educational).
The particular rumor of NV entering the CPU market goes back several years. as an example
Re:Okay, I'll bite... (Score:5, Informative)
actually, it was IBM who required two sources (per their own company policy)
later, legal disputes settled the question
Parent
Re:Okay, I'll bite... (Score:5, Interesting)
> I don't see why nVidia would have it any different or wouldn't be able to do the same.
There are already a crapload of people who are now or have in the past sold x86 compatible chips. You generally can't claim an exclusive on the public interfaces like the instruction set. Patents on various sub systems are a problem for anyone doing anything these days, but new products somehow manage to get to market.
Cyrix->Via didn't have a cross license deal. Don't think Transmeta did either. I even remember 8086 compatible chips with NEC's stamp on the package. And there are a couple more I remember existed but can't pull the name from memory. The skills to make an x86 compatible processor from scratch is getting pretty widespread, making and selling one competitive with Intel and AMD is a different kettle of fish as so many others have found to this dismay.
Parent
Re:Okay, I'll bite... (Score:5, Informative)
Via's current chips are descendants of the IDT WinChip. Basically, a design similar to the original Pentium, IIRC.
From memory, and I may be missing some, here's all of the unlicensed x86 compatibles and their descendants:
NEC V20/V30 (these may be licensed, I forget whether NEC was a second source for Intel, as well)
Cyrix 486DLC/486SLC/486/5x86/MediaGX (now sold as the AMD Geode GX1)/6x86/M2 - this line was going to evolve into the VIA Cyrix III, but the 3rd-generation Centaur design, which was supposed to be the low power/budget chip, was also much faster
IDT WinChip/WinChip 2/VIA C3/VIA C7 - there's a modern descendant to this line that I forgot the name of, but can FINALLY do out of order execution
NexGen Nx586 - the successor to this was almost the Nx686, but AMD bought NexGen, repackaged it, and called it the K6
IBM Blue Lightning 486 - yes, IBM did their own 486 design, while also manufacturing Cyrix 486s
Rise mP6/SiS550
The VIA chips are proof that licensing isn't necessary - they support 3DNow, MMX, and various forms of SSE...
Parent
TFA is nearly as useless as the summary (Score:3, Informative)
I was going to tell you to RTFA but TFA is almost as useless as the summary. Apparently Intel and AMD have a "lock" on the technology. What part of the technology they have a "lock" on is left unsaid... the instruction set? The manufacturing processes? TFA doesn't bother to say.
Re:TFA is nearly as useless as the summary (Score:4, Informative)
I was going to tell you to RTFA but TFA is almost as useless as the summary. Apparently Intel and AMD have a "lock" on the technology. What part of the technology they have a "lock" on is left unsaid... the instruction set? The manufacturing processes? TFA doesn't bother to say.
Intel and AMD have a lock on all the instruction sets that makes modern processors "modern".
SSE, MMX, 3DNOW!, AMD64, Intel 64, etc are all cross licensed between the two companies.
If they don't want to share, there isn't much anyone else can do.
Parent
Re:TFA is nearly as useless as the summary (Score:4, Interesting)
AFAIK, the VIA Nano supports all of those four instruction sets (Intel 64 and AMD64 are basically the same ISA). None of them are specific to Intel or AMD anymore.
Parent
Re:Okay, I'll bite... (Score:5, Informative)
Socket and interface patents. Intel have patents on various bits of the interface between the CPU and the motherboard, which is one of the reasons why AMD use a different one for their CPUs.
Assuming nVidia is going to make a pin-compatible processor with one of the motherboard sockets already out there, they'll need a licence from intel or AMD. That's assuming they don't produce a small low power chip wedded to a particular board, like say the intel atom or the via nano, aiming for the new netbook market or the mini pc segment.
As I understand it, they already had to cough up a SLI licence to intel in order to get a licence from intel to make nehalem compatible motherboard chipsets, which means we'll finally see realistic motherboards with sli and crossfire.
Parent
Re: (Score:3, Insightful)
Re:Okay, I'll bite... (Score:5, Informative)
Um no. More like long ago Intel needed a second source for CPU's and contracted AMD, granting a license in the process. AMD didn't start making any superior advances in CPU design for quite some time after that.
Now...get off my lawn.
Parent
Re:Okay, I'll bite... (Score:5, Insightful)
So, unless a new player comes up with some amazing new technology which Intel and AMD want enough to let that company into the patent-party, it would be very difficult for a new competitor to walk in.
What are the odds of nVidia finding some patent violation in Intel's or AMD/ATI's graphics chips? Would nVidia be able to play the Mutually Assured Destruction card?
Parent
Re: (Score:3, Informative)
Re:Okay, I'll bite... (Score:5, Insightful)
Parent
Re: (Score:3, Insightful)
The original article is atrocious. There are no details on what type of licensing this fellow is claiming and he throws in some patent leverage that allows VIA to get away with making x86 parts.
The author of the original article sounds like he doesn't have a clue what he has heard and has no idea how to explain it.
From what I can tell from his badly munged writing, it looks like nVidia can make x86 compatible processors to take on Intel and AMD in the performance processor market, but they are in a legal bi
Re:Okay, I'll bite... (Score:4, Informative)
How is it that AMD is able to release x86 chips, but nVidia can't without a license from Intel? Why would nVidia need AMD to be gracious?
According to wikipedia nVidia does have license from its purchase of what was left over of 3dfx. Remember them?
Parent
Re:Okay, I'll bite... (Score:5, Interesting)
As far as lawsuits go, I can't see Intel opening up on nVidia either way. They already have enough problems in the US and EU with anti-trust threats: a lawsuit against a new player would be just be more evidence against them. Secondly, nVidia are a major producer of chipsets for Intel, and a lawsuit could see them dump support for Intel (and either solely support AMD or leave the chipset business altogether).
nVidia haven't been faring too well lately, and entering into this venture would have been well researched. I imagine that patents and licensing would have been one of the first issues to get sorted.
Personally, I don't think nVidia would be capable of entering the market with x86 chips for high-end desktops. However, I do suspect that we'll see a system-on-a-chip based on Via's Nano, with an on-chip GPU (with PhysX and SLI capability), northbridge and southbridge. It would be rather like nVidia's Tegra is to ARM11. I'm guessing that such a processor wouldn't be destined for the desktop, but rather the portable market.
Parent
What about VIA? (Score:5, Interesting)
Does VIA has a license to make x86 processors?
Re:What about VIA? (Score:5, Informative)
On the basis of the IDT Centaur acquisition,[1] VIA appears to have come into possession of at least three patents, which cover key aspects of processor technology used by Intel. On the basis of the negotiating leverage these patents offered, in 2003 VIA arrived at an agreement with Intel that allowed for a ten year patent cross license, enabling VIA to continue to design and manufacture x86 compatible CPUs. VIA was also granted a three year grace period in which it could continue to use Intel socket infrastructure.
So the answer to your question is: Yes, but only until 2013.
Parent
Re: (Score:3, Informative)
VIA actually bought Centaur [wikipedia.org] that had patents on x86 manufacturing. But intel sued VIA and VIA reciprocated in what amounted to be a long and protracted litigation. Eventually they settled [via.com.tw] after a judge ordered them to do so (as I suspect that it was too much of a technical mess for most judges to wade through).
I actually wondered what VIA would be able to do without being able to produce a pin compatible x86 processor. But that would be answered with the very unique mini-itx line of boards which is differen
Emulation/Translation - do it in software? (Score:5, Interesting)
They could pull a Transmeta and build a RISC/VLIW core or six and package it with an x86 interpreter or JIT translator, basically do the front end in software instead of hardware. Crusoe was using the same core to do the translation and execution, but with a multi-core CPU that pipelines the translator and interpreter on separate cores they could end up with quite a nice design.
Nvidia would not need a license everywhere! (Score:5, Interesting)
Sure, if Nvidia tried selling x86 chips in the US or Europe, the company would get its ass sued off. But what about China? What about India? What about the third world? Merely because Intel has a rock solid patent portfolio in the US does not mean diddly squat in Bangladesh.
Re: (Score:3, Insightful)
Sure they do.
Nvidia has a company presence in the US so they can get sued here.
Not only that but India and China want to do business with Intel and AMD so they will not be real happy with open disregard for IP.
They may be perfectly happy to ignore IP but they know they must be subtle about and at least pretend to fight it.
Re: (Score:3, Insightful)
x86 is basically free for the taking. MMX, SSE, SSE2, SSE3, 3dNow!, and x86-64 are not.
Must have a Legal Plan before starting... (Score:5, Insightful)
Given the cost of developing a full custom microprocessor (several tens of millions of dollars) including the complexity of verification ... surely a Legal Plan would have proceeded either development or acquisition.
Re: (Score:3, Insightful)
There is no legal problem if they reverse enginneer it and don't copy the socket design. There is not a single legal barrier to making a processor that can decode x86 instructions.
Don't worry about the licensing. (Score:5, Insightful)
interesting (Score:3, Insightful)
though currently, these are only rumors, it would be interesting to see how it will play out if these turn out not to be rumors.
For one, aren't both Intel and AMD having their own problems with anti-trust litigation in various places around the world? (I know Intel and the EU like to go at it)
Intel might just quickly license nVIDIA to do so just so that they can claim that there is no anti-trust going on, especially when there's a 3rd player at the table.
How quickly we forget... (Score:5, Interesting)
Cyrix [wikipedia.org], Texas Instruments [wikipedia.org], IBM [wikipedia.org], NexGen [wikipedia.org], amongst others.
Other companies made clone x86 [wikipedia.org] CPUs as well (The list: IBM, NEC, AMD, TI, STM, Fujitsu, OKI, Siemens, Cyrix, Intersil, C&T, NexGen, and UMC). Intel has never been really successful at prosecuting anyone for creating their own x86 compatible CPU. They won't sue, unless the company is small enough to just give up (Hint: nVidia isn't).
Interesting legal histories (Score:5, Informative)
Check out the legal histories of AMD v. Intel and VIA/Cyrix v. Intel. These essentially show that there are agreements and settlements all over the place, but few-to-no actual court decisions.
http://en.wikipedia.org/wiki/VIA_Technologies#Legal_issues [wikipedia.org]
http://en.wikipedia.org/wiki/Cyrix#Legal_troubles [wikipedia.org]
http://en.wikipedia.org/wiki/AMD#Litigation_with_Intel [wikipedia.org]
It essentially seems that NVIDIA would need to have a patent on something which Intel has produced in order to induce some kind of Mexican standoff, just like the others have.
Something I forgot to mention in the summary (Score:5, Interesting)
jdb2
A "license" for unrelease, unannounced rumors (Score:4, Insightful)
So can someone tell me how anyone outside of NVidia (who isn't quoted here) would know they need a "license" (patents I'm assuming) for a technology that nobody knows anything about, is completely unreleased, and likely doesn't even exist?
This story is complete nonsense. We're all dumber after having read it.
Re:A "license" for unrelease, unannounced rumors (Score:5, Funny)
Parent
half right (Score:3, Interesting)
Look guys, here's the facts: Nvidia will be releasing a chip. It won't be x86, though, it will be ARM based (with an fpu and vector unit), running around 1GHZ or more. A couple months ago, ARM Holdings announced a major license agreement (but didn't provide any other specifics). There was a lot of speculation that it was Apple. It's Nvidia.
My source didn't tell me if it's going to be targetted at smart phones, internet tablet pcs, etc.
Old news... (Score:5, Informative)
Cyrix did it. (Score:5, Informative)
Cyrix originally didn't license anything. They reverse engineered 386/486 designs. Intel sued them over it and mostly lost. The settlement allowed Cyrix to continue producing the designs, provided they were made in Intel licensed factories. Later, Cyrix nailed Intel infringing on some of their patents, and it was settled by allowing each to use the others patents.
If Nvidia tries to produce their own CPU, I would guess they'd be sued, but it would probably end in a pro-nvidia settlement. I suspect Nvidia holds some patents they can dangle over Intel's head.
Anyways, all of the speculation is meaningless, if Nvidia is actually doing this they've got the legal parts taken care of.
Implement "Not X86" X86 (Score:4, Interesting)
Seems easy to bypass the X86 issue. Create a full CPU using the X86 instruction set. Remove anything and everything related to 286 protected mode (keep "real" and 386 "protected" modes). Optionally, remove ring 1 & 2 from 386 protected mode, but keep the register format the same (windows and unix only use ring 0 and ring 3). Then, add a new CPU instruction or two that would really boost the performance of Nvidia's graphics drivers, which Nvidia can autodetect and use in their shipping drivers (just like most graphics drivers used to detect SSE and the like). Naturally, no one else would use these instructions, but Nvidia could be a good citizen and document them.
The resulting chip wouldn't be X86, because all X86 code does not run. The result would be a new chip that isn't backwards compatible. Let Intel bark and moan all day long in their marketing that the chip isn't X86. All Nvidia has to do is make sure it runs Windows just fine without a new SKU from Microsoft (is it Intel's fault that MS doesn't use 286 protected mode? Is it Intel's fault that MS doesn't use ring 1 or ring 2?).
There would still be a lawsuit, and it would be *wise* to ensure that your legal team is well funded. But it seems most legal arguments are letter of the law these days, and the subset and extended X86 is definitely not X86 (you can produce code that works on X86, but would fail on this, you can produce code that works on this but fails on X86).
This would be a ballsy move for Nvidia, but seems right up their alley.
Forget the license, what about fabs? (Score:4, Insightful)
I assume Nvidia has some juicy tech they could cross-license to Intel and AMD in return for the rights to make their own x86.
But who will build it? Last time I checked Nvidia didn't own a fab plant. All their stuff is built by TSMC, a very respectable GPU fab but still a generation behind Intel in process technology. Unless Nvidia has some secret fab project going for the last ten years, they certainly don't have "guns targeted directly" at Intel or AMD.
Now if you told me they were going to compete with VIA in the ultra-low-power SOC [wikipedia.org] market, that might be interesting. Still, I imagine Nvidia has better things to do than throw resources at such a low-margin business.
Re: (Score:3, Informative)
Re:A question about x86 versus RISC (Score:4, Interesting)
Given legal and licensing issues, it makes sense to work around this issue with a RISC / VLIW core (and NVidia has already mastered this) with a JIT or x86 bytecode interpreter at the front end.
Intel (and everyone else) has been doing this, to an increasing degree, since the 486. It's one reason for the long long Pentium pipeline.
Transmeta made the JIT translator software. That turned out to be not such a good idea, with one core trying to execute translated code and translate new code at the same time. With multiple cores (and nVidia has a lot of experience with heavy parallelism like that) this could actually work well.
My comment here [slashdot.org] was idle speculation along these lines. But apparently nVidia has been speculating less idly: they've licensed the Transmeta technology.
Parent