Slashdot Mirror
Nvidia Is Trying To Make an x86 Chip
Slatterz writes with a story from PC Authority which says that "Word has reached us that Nvidia is definitely
working on an x86 chip and the firm is heavily recruiting x86 engineers all over Silicon Valley. The history behind this can be summarised by saying they bought an x86 team, and don't have a licence to make the parts. Given that the firm burned about every bridge imaginable with the two companies who can give them licences, Nvidia has about a zero chance of getting one."
Nvidia are going to challenge the concept of licensing an instruction set, and they know they are going to win.
That will be a great day for all the technology industry and herald a massive price crash in processor power.
The day after he brought you news about Intel creating the Playstation 4 GPU discussed here comes more industry shaking news, original article here.
Wow, that's two pretty big news scoops on back to back days for Charlie with both making Slashdot's homepage at the same time!
Would those patents include an in hardware x86 instructionset translator to their GPU instructionset? I remember some vague comments around ~5 years ago that nVidia wanted to run an OS on their GPUs.
Support Eachother, Copy Dutch Property!
Shame about that, at least try and find some additional information and link to the original article. I didn't know that the INQ has become a news agency of sorts. They certainly don't have the credentials for that. And the author of this article even less.
Then again, we can discuss the idea that nVidia is apparently (no proof whatsoever of the hirings) going for x86 without having the licenses to do so. As I understood, AMD and Intel (and VIA) let each other use patents and designs for x86, so I assume this is about letting nVidia in or not on that scheme.
Personally I'm wondering why nVidia and VIA don't fuse. One just has created a neat little x86 CPU and low power parts the other has neat GPU's. And I heard that VIA is going out of the chipset business anyways.
See, I've started up the discussion for you. If you don't like it you can order up another if you don't think it's any good.
When I saw the summery this is the 1st thing that came to my mind.
What is all they want to do is use the high density chip technology they currently have to produce a 3 Ghz or faster 80386DX CPU ?
One with all the RAM it can handle as (core speed) cache?
--= Isn't it surprising how badly I spell ?
Which leads to in important consideration. Yes, AMD and Intel hold patents vital to getting into the industry, but why did they cross license? That is simple: Intel had enough patents to have AMD by the balls, and AMD had enough patents to have Intel by the balls. Neither enjoyed being at the mercy of the other, so they came to a mutual agreement.
So now, fast forward to present day. Nvidia wants to get into the game. So how do they do it? Simple: they need to innovate and get patents on core technology before the other 2 do. Then they can agree to license it to one of the 2 to give them a competitive advantage. At that point Nvidia has half the necessary portfolio, and if things go well, the other will need to get their hands on the tech to stay competitive.
um there has been massive amounts of changes to the x86 design line over the last 20 years too.
To the point where they are almost superficially x86.
i thought once I was found, but it was only a dream.
Maybe Nvidia should talk to Motorola instead.
I'd love to see a modern-day version of the 68060.
"I disapprove of what you say, but I will defend to the death your right to say it." - historian Evelyn Beatrice Hall
So why not just buy Via? They have the license to make x86, and more importantly they have low power CPUs that are ready to go, and with Netbooks and Notebooks taking a big chunk out of the market this would give them a BIG advantage in the market. If they were to buy Via they could join the Via Nano and the Ion GPU and have bad ass low power Netbooks and Notebooks ready to hit the market.
So can someone tell me what the point of doing it the hard way by starting from scratch is? Because IMHO it seems like a giant waste of resources when Via is ripe for the picking and the Nano CPU from what I have seen is a nice mix of low power with decent performance for Netbooks/Notebooks. Paired with the Ion GPU I think they would have a combo that would kick some serious ass in the Netbook and Notebook markets with little effort. Then later on if they desired they could always do a Fusion style joining of the CPU/GPU to get even better power to performance ratio. So why are they trying to reinvent the wheel?
ACs don't waste your time replying, your posts are never seen by me.
Cross-licensing is a crock. It is done to try and head off any threat of legal action two or more companies might throw at each other, but the suspicion of that threat is not based on anything concrete. It's more about warm fuzzy feelings and to give the legal people something to do. It's also done as a protectionist tactic between companies to make sure no one else enters the party, and if they try to to ensure that everyone will be asking a lot of questions that can't be answered about their legality.
I've torn apart quite a few 286 and 386 computers where the chip clearly says "Intel 286" or "Intel 386SX" and in small print down by the serial # for the processor, it is stamped "AMD" or "Made by AMD" or something along those lines (I can't exactly remember how the AMD part was worded)
Just 4 Gb of RAM, a 32-bit address, and make it as fast as you can. Forget about that 64-bit bullshit, I'm not running the Social Security database. But it must be on a single chip, or as close as it can be. Memory access times are limited by the speed of light once you get into the GHz range, a nanosecond is 300 millimeters.
To go with that, let's have some thousands of cores for number crunching. Mega cores, giga cores, you can never have enough cores for number crunching. But these cores need not have 64-bit capability, all they need to do is multiply-add operations, as quickly as possible.
The CPU industry, unfortunately, has been too long in a monopoly situation. Nvidia has done some impressive progress in getting an alternative thinking to the market, let's see what they can bring next.
Could be.
There was a big fight in the chipmaking world over a bunch of patents covering hardware x86/Instruction set translation, which included multicore parallel instruction processing. They were originally held by a company called Exponential Technologies, and though Intel wanted them badly, were grabbed by S3 for ten million in an auction.
In the end, S3 and Intel agreed on a time-limited cross licensing deal. That agreement ended in December 2008.
Coincidence?
"I've got more toys than Teruhisa Kitahara."
Since S3 Graphics = Via... I think buying stock sounds attractive.
Support Eachother, Copy Dutch Property!
The point is that you could make a totally different core chip and program the x86 instruction set to run on it. nVidia already has fast chips and motherboards. I think they are close to selling the whole package. Imagine a micro-ATX single-board system where you aren't paying for both a CPU and GPU. Small size and power, low cost and high performance.
Ancient history. AMD got into the x86 market in the 80's when the USG required multiple sources for many components
You know, in hindsight, keeping up this policy would have been a vastly more effective way of reigning in Microsoft than that ineffectual antitrust suit.
Nobody building a contemporary x86 CPU "starts" with x86 as it is. Much of the "x86" aspect of our processors is little more than an abstraction layer, translating to the functional guts of the CPU. The guts are designed with a mind for the needs of that abstraction layer, but it's not like the old days, where a certain x86 instruction mapped to a hard-wired set of transistors on the die.
If a job's not worth doing, it's not worth doing right.
This is because, in the good old days, AMD got their start by being an official second source for official Intel chips. They were allowed to build, up through the 386, the exact same chip that Intel was building. Intel effectively gave them their start by being the "2nd source" because in those days, many companies would not consider using Intel chips until they were assured of a second source for the same chips. This was, of course, before Intel became nearly the only game in town.
Agreed... sometimes starting from scratch is easier than fixing old code. But IF they could buy Via, it wouldn't mean they HAVE to use the existing code-base. They could buy Via just for the licenses and build from scratch anyway.
-- I ignore anonymous replies to my comments and postings.
Yes, but the patents on the (generic) x86 instruction set are all expired, considering that the 80286 is more than 20 years old. So while Intel (32-bits) and AMD (64-bits) may hold a number of patents on the design of those chips, NVidia might be able to pull a Transmeta on them (ie converting the x86 instruction set to their own execution format as part of the Instruction Decode phase) without bending over for royalties.
If they play their cards right and get to market first, they might even set the standard for future x86 MIMD instructions. That would allow them to collect royalties from both the other players, much like AMD beat Intel on getting a workable 64-bit implementation.
But I think the odds are 90% that NVidia is screwed, unless they get a judicial clearance first (like a monopoly case against Intel).
Trivia: The turbo button/LED was actually to slow down the PC for clock cycle-based applications that ran too fast on modern PCs. See MoSlo for a software implementation. ;)
Homonyms are fun!
You're driving your car, but they're riding their bikes there.
Yes, but the patents on the (generic) x86 instruction set are all expired, considering that the 80286 is more than 20 years old. So while Intel (32-bits) and AMD (64-bits) may hold a number of patents on the design of those chips, NVidia might be able to pull a Transmeta on them (ie converting the x86 instruction set to their own execution format as part of the Instruction Decode phase) without bending over for royalties.
If they play their cards right and get to market first, they might even set the standard for future x86 MIMD instructions. That would allow them to collect royalties from both the other players, much like AMD beat Intel on getting a workable 64-bit implementation.
But I think the odds are 90% that NVidia is screwed, unless they get a judicial clearance first (like a monopoly case against Intel).
I'm thinking the same thing, with a twist. Instead of creating a CPU, they might be trying to build a Transmeta-like converter to work with their GPU technology to support Larabee-like extensions to the x86 instruction set. Perhaps NVidia knows that the future is in hygrid ray-tracing with rasterization, and one of the ways to get developers to keep supporting the NVidia platform is to make NVidia's platform support the same crap that Intel is devising with Larabee. That way, game and graphics devs don't need to work as hard trying to fit Larabee ray-tracing techniques into a rasterization engine on the GeForce platform.
Thoughts?