Slashdot Mirror
Nvidia CEO "Not Afraid" of CPU-GPU Hybrids
J. Dzhugashvili writes "Is Nvidia worried about the advent of both CPUs with graphics processor cores and Larrabee, Intel's future discrete graphics processor? Judging by the tone adopted by Nvidia's CEO during a financial analyst conference yesterday, not quite. Huang believes CPU-GPU hybrids will be no different (and just as slow) as today's integrated graphics chipsets, and he thinks people will still pay for faster Nvidia GPUs. Regarding Larrabee, Huang says Nvidia is going to 'open a can of whoop-ass' on Intel, and that Intel's strategy of reinventing the wheel by ignoring years of graphics architecture R&D is fundamentally flawed. Nvidia also has some new hotness in the pipeline, such as its APX 2500 system-on-a-chip for handhelds and a new platform for VIA processors."
Did I hear that correctly? NVidia is going to beat Intel in the GPU department? What a breaking development!
In other news, Aston Martin makes better cars than Hyundai!
No competition? What? Did ATI die or something?
Yes I know they got bought by AMD, but they still exist and they still make GPUs AFAIK.
And if your argument is that nVidia is better than ATI, let me remind you that ATI/nVidia and intel/AMD keep leapfrogging each other every few years.
What AMD should really try to do is start combining their cpu technology and their graphics technology and make some multi core GPUs. They might be better positioned to do this than Intel or Nvidia.
Until Intel can show us Crysis in all it's GPU raping glory running on it's chipset in 1600x1200 with all settings to Ultra High Nvidia and ATI will still be kings of high end graphics. Then again, if all Intel wants to do is create a sub standard alternative to those high end cards just to run Vista Aero and *nix Beryl then they have already succeeded.
IMHO, Nvidia is stuck as the odd-man out. When integrated chipsets and GPU-CPU hybrids can easily handle full-HD playback, the market for discrete GPUs falls and falls some more. Sure, discrete will always be faster, just like a Porsche is faster than a Toyota, but who makes more money (by a mile)?
Is Creative still around? Last I heard, they were making MP3 players...
I, for one, don't want a GPU which requires 25W+ in standby mode.
My Mac mini has a maximum load of 110W. That's the Core 2 Duo CPU, the integrated GMA950, 3GB of RAM, a 2.5" drive and a DVD burner, not to mention FireWire 400 and four USB 2.0 ports under maximum load (the FW400 port being 8W alone).
Granted the GMA950 sucks compared to nVidia's current offerings, however do they have any plans for low-power GPUs? I'm pretty sure the whole company can't survive on the FPS-crazed game players revenues alone.
They should start thinking about asking intel to integrate their (current) laptop GPUs into intel CPUs.
If I understand them right, they're claiming that integrated graphics and CPU/GPU hybrids are just a toy, and that you want discrete graphics if you're serious. Ken Olsen famously said that "the PC is just a toy". When did you last use a "real" computer?
Ray vs raster. The reason we have so much tech in Raster is because processing was not sufficient to do ray. If it had been we'd have never started down the raster branch of development because it just doesn't work as well. The results are not as realistic with raster. Shadows don't look right. You can't do csg. You get edge effects. There are a thousand work-arounds for things like reflections of reflections, lens effects and audio reflections. Raster is a hack and when we have the CPU to do the real time ray tracing rendering raster composition will go away.
Raster was a way to make some fairly believable (if cartoonish) video games. They still require some deliberate suspension-of-disbelief. Only with raytracing do you get the surreal Live-or-memorex feeling of not being able to tell a rendered scene from a photo, except for the fact that the realistic scene depicts something that might be physically impossible.
Help stamp out iliturcy.
If Intel is right, there won't be much of an effect on existing games.
Intel is focusing on raytracers, something Crytek has specifically said that they will not do. Therefore, both Crysis and any sequels won't really see any improvement from Intel's approach.
If Intel is right, what we are talking about is the Crysis-killer -- a game that looks and plays much better than Crysis (and maybe with a plot that doesn't completely suck), and only on Intel hardware, not on nVidia.
Oh, and Beryl has been killed and merged. It's just Compiz now, and Compiz Fusion if you need more.
Don't thank God, thank a doctor!
"Huang says Nvidia is going to 'open a can of whoop-ass' on Intel..."
This is a VERY SERIOUS problem for the entire world. There are apparently no people available who have both technical understanding and social sophistication.
Huang is obviously ethnic Chinese. It is likely he is imitating something he heard in a movie or TV show. He certainly did not realize that only ignorant angry people use that phrase.
Translating, that phrase, and the boasting in general, says to me: "Huang must be fired. nVidia needs a better top manager."
ATI/AMD hasn't been competitive with NVIDIA for two product cycles. That doesn't look likely to change in the near future, either; ATI/AMD's next generation GPU architecture isn't looking so hot.
AMD is in a world of hurt right now, with Intel consistently maintaining a lead over them in the CPU segment, and NVIDIA maintaining a lead over them in the GPU segment. They're doing some interesting, synergistic things between the CPU and GPU sides, but who knows if that'll pan out. Meanwhile, they're being forced to compete on price alone, which is never a position you want to be in.
(The driver quality situation hasn't exactly helped them any, either, although I'm looking forward to good things post-acquisition, especially now that open source drivers are becoming a reality.)
NVDA was down 7% in the stock market today. As an Nvidia shareholder, that hurts!
If you don't believe Intel will ever compete with Nvidia, now is probably a good time to buy. NVDA has a forward P/E of 14. That's a "value stock" price for a leading tech company... you don't get opportunities like that often. NVDA also has no debt on the books, so the credit crunch does not directly affect them.
A slashdotter who didn't build his own computer is like a Jedi who didn't build his own lightsaber.
Intel is and always has been CPU-centric. That's all they ever seem to focus on because it's what they do best. Nvidia is focusing 100% on GPUs because it's what they best. AMD seems to have it right with their combination of the two (by necessity) because they're focusing on a mix between the two. I'm seriously stoked about the 780G chipset they rolled out this month because it's an integrated chipset that doesn't suck and actually speeds up an ATI video card if you add the right one. Given, AMD isn't the fastest when it comes to either graphics or processors but at least they have a platform with a chipset, CPU, and graphics that work together. Chipsets have needed to be a bit more powerful for a long-ass time.
"Bus and train integration is quite logical progression of technology. There are things the plane is not optimal and same goes to the bus. It seems that when combined, they prove successful. So let's put wings on a bus."
Now, I think there are plenty of good reasons why CPU/GPU integration is a good idea (as well as a few good reasons why it's not), but there's nothing logical about the statement you made. Just because a CPU does something well and a GPU does something different well, it doesn't necessarily follow that slapping them together is a better idea than having them be discrete components.
The key insight is that the modern CPU and the modern GPU are starting to converge in a lot of areas of functionality. The main difference is that CPUs are optimized for serial processing of at most a few threads of arbitrarily complex software, while GPUs are optimized for massively parallel processing of large numbers of pixels using similar, fairly simple programs (shaders).
Now, the logic core needed to perform these two tasks is highly specific, which is why we have separate CPUs and GPUs to begin with. But there's a lot to be gained by integrating the two more closely. You can share memory interfaces, for example, and perhaps more relevantly for the high-end graphics segment, you can tightly couple CPU and GPU operations across a bus that's going to be a hundred times faster than anything PCI Express can provide, and with latency to die for.
In short, I agree with your basic point, but I don't think you made a very good case for it.
> ATI/AMD hasn't been competitive with NVIDIA for two product cycles
Competitive enough anyway. Long as I'm still on AGP, I'm still getting ATI cards (nVidia's agp offerings have classically been highly crippled beyond just running on AGP). But sure, I'm a niche, and truth be told, my next system will probably have nVidia.
But gamer video cards aren't everything, and I daresay not even the majority. If you have a flatscreen TV, chances are good it's got ATI parts in it. Then there's laptops and integrated video, nothing to sneeze at.
Done with slashdot, done with nerds, getting a life.
Granted NVidia is way out ahead in graphics performance - but generally you can tell when that someone is getting nervous when they start in the belligerant bragging.
The risk for NVidia isn't that Intel will surpass them, or even necessarily approach their best performance. The risk is that Intel might start catching up, cutting (further) into NVidia's market share.
AMD's acquisition of ATI seems to imply that they see tight integration of graphics to be at least cheaper for a given level of performance, or higher performance for a given price. Apply that same reasoning to Intel, since they certainly aren't likely to let AMD have that advantage all to themselves.
Now try to apply that logic to NVidia - what are they going to do, merge with a distant-last-place x86 maker?
Of COURSE they do, in fact they already HAVE low power offerings. I'm not sure why people seem to think the 8800 is the only card nVidia makes. nVidia is quite adept at taking their technology and scaling it down. Just reduce the clock speed, cut off shader units and such, there you go. In the 8 series they have an 8400. I don't know what the power draw is, but it doesn't have any extra power connectors so it is under 75 watts peak by definition (that's all PCIe can handle). They have even lower power cards in other lines, and integrated on the motherboard.
So they already HAVE low power GPUs. However you can't have both low power and super high performance. If you want something that performs like an 8800, well, you need an 8800.
Once upon a time the floating point was done on a seperate chip. You could buy a cheaper "non-professional" machine that emulated the fpu in software and ran slower. You could also upgrade your machine by adding the fpu chip.
Such FPU's do not exist today.
I think Nvidia should be worried about this.
...is that Nvidia is saying that Intel is ignoring years of GPU development. Umm, wait. Isn't a GPU basically a mini-computer/CPU by itself that exclusively handles graphics calculations? By making this statement, I think they've forgotten who Intel is. Intel has more than enough experience in the field to go off on their own and make GPUs. Is it something to be scared of? Probably not, because as he correctly points out, a dedicated GPU will be more powerful. However, it's not something that can be ignored. We'll just have to wait and see.
The Computations of AdamR
http://www.adamreyher.com
This is a VERY SERIOUS problem for the entire world. There are apparently no people available who have both technical understanding and social sophistication.
Maybe he was out of chairs?
Next you're going to tell me the sky is blue or that too much water can kill me. Onboard video isn't meant to be shiny, just to serve a basic need: being able to see what the hell you're doing. Rather than dismissing Intel because they (and many other board manufacturers) provide a bare-bones video solution, I'm interested in seeing what they'll pop out when they're actually trying.
By the way, onboard video uses about as much RAM as a browser will use (And about as much as Win98 needs to boot in, but I digress.), hardly a drop in the bucket with 1GB sticks being so cheap now. If 8-32MB of RAM is that much of a problem for you, you have more problems than poor video.
I just read Slashdot for the articles.
Huang is obviously ethnic Chinese. It is likely he is imitating something he heard in a movie or TV show.
Yeah, them slanty-eyed furriners just can't speak English right, can they?
Huang is over 40 years old and has lived in the US since he was a child. Idiot.
Done with slashdot, done with nerds, getting a life.
Quite true. With the 8500 and 8600 models, and now the 9500, nVidia trounces AMD even on budget cards.
But, nVidia got pummeled prior to their acquisition of Yahoo!^H^H^H^H^H^H Voodoo, and the two were quite neck and neck for a long time. So it's more of "the tables have turned (again)" rather than "they have no competition."
Until AMD completely quits making higher-end video cards, nVidia will have to keep on doing something to stay competitive. Same thing with Firefox - I don't think IE8 would have looked any different than IE5 without something biting at their heels-slash-completely surpassing them.
DATABASE WOW WOW
Having the GPU built into the CPU is primarily a cost-cutting measure. Take one low-end CPU, add one low-end GPU, and you have a single-chip solution that consumes a bit less power than separate components.
Nobody expects the CPU+GPU to yield gaming performance worth a damn, because the two big companies that are looking into this amalgam both have underperforming graphics technology. Do they both make excellent budget solutions ? Yes they certainly do, but for those who crave extreme speed, the only option is NVidia.
That said, not everyone plays shooters. Back in my retail days, I'd say I moved 50 times more bottom-end GPUs than top-end ones. Those Radeon 9250s were $29.99 piles of alien poop, but cheap poop is enough for the average norm. The only people who spent more than $100 on a video card were teenagers and comic book guys (and of course, my awesome self).
-Billco, Fnarg.com
The problem is not that Nvidia CEO Jen-Hsun Huang made one stupid statement. The problem is that he said many foolish things, indicating that he is not a good CEO. Here are some:
Quote from the article: "Nvidia CEO Jen-Hsun Huang was quite vocal on those fronts, arguing hybrid chips that mix microprocessor and graphics processor cores will be no different from systems that include Intel or AMD integrated graphics today."
My opinion: There would be no need for all the talk if there were no chance of competition. Everyone knows there will be new competition from Intel Larabee and AMD/ATI. Everyone knows that "no different" is a lie. Lying exposes the Nvidia CEO as a weak man.
"... he explained that Nvidia is continuously reinventing itself and that it will be two architectural refreshes beyond the current generation of chips before Larrabee launches."
The entire issue is that Intel+Larabee and AMD+ATI will make Nvidia irrelevant for most users. The GPU will be on the motherboard. Nvidia will sell only to gamers who are willing to pay extra, a lot extra.
"Huang also raised the prospect of application and API-level compatibility problems with Larrabee. Intel has said Larrabee will support the DirectX 10 and OpenGL application programming interfaces just like current AMD and Nvidia GPUs, but Huang seemed dubious Intel could deliver on that front."
Intel, in this case, is Intel and Microsoft working together. Both are poorly managed companies in many ways, but they are both managed well enough to insure that the Microsoft product works with the Intel hardware. Sure, it is an easy guess that Microsoft will release several buggy versions, because Microsoft has a history of treating its customers as though they were beta testers, but eventually everything will work correctly.
'[NVidia VP] Tamasi went on to shoot down Intel's emphasis on ray tracing, which the chipmaker has called "the future for games." '
Ray tracing is certainly the future for games, there is no question about that. The question is when, because the processor power required is huge. It's my guess, but an easy guess, that Mr. Tamasi is lying; he is apparently trying to take advantage of the ignorance of financial analists.
"Additionally, Tamasi believes rasterization is inherently more scalable than ray tracing. He said running a ray tracer on a cell phone is "hard to conceive."
This is apparently another attempt to confuse the financial analyists, who often have only a pretend interest in technical things. Anyone understanding the statement knows it is nonsense. No one is suggesting that there will be ray-tracing on cell phones. My opinion is that this is another lie.
"We're gonna be highly focused on bringing a great experience to people who care about it," he explained, adding that Nvidia hardware simply isn't for everyone."
That was a foolish thing to say. That's the whole issue! In the future, Nvidia's sales will drop because "Nvidia hardware simply isn't for everyone." Most computers will not have separate video adapters, whereas they did before. Only powerful game machines will need to by from Nvidia.
'Huang added, "I would build CPUs if I could change the world [in doing so]." ' Later in the article, it says, "Nvidia is readying a platform to accompany VIA's next-generation Isaiah processor, which should fight it out with Intel's Atom in the low-cost notebook and desktop arena"
Translation: Before, every desktop computer needed a video adapter, which came from a company different than the CPU maker, a company like Nvidia. Now, the video adapters will be mostly supplied by CPU makers. In response, Nvidia will start making low-end CPUs. It is questionable whether Nvidia can compete with Intel and AMD making any kind of CPU.
I thought it was DirectX 9 that they were left out from, causing their FX range (5200 -> 5900) to be fairly useless when compared to ATI. They were legends in the DirectX 8 arena with the GeForce4 Ti series.
I haven't, but I have heard of Carmack, and Carmack "seems to think that Intel's direction using traditional ray tracing methods is not going to work." I didn't understand anything in that article, but assuming that the blurb was correct (and Carmack didn't seem to refute it in the 3 times he replied to that story), then I'd say that they may not be "less and less interested" but maybe they are "less and less right about the direction to take." Take your pick.
And while my little blurb may have been fundamentally incorrect, while I haven't heard anyone say they were looking forward to the new Radeon, I have heard even less people were looking forward to the new Intel graphical chipset. Have you?
Splitting hairs on this seems kinda useless. nVidia is really it right now in the graphics world, at least as far as the public is concerned, and I don't see that changing in the near future.
The large corporations and engineering companies that have *THOUSANDS* of high-end workstations need graphics hardware compatible with complex, specialized software. I'm talking Unigraphics, CATIA, Patran, Femap, etc. You need to use the hardware certified by the software publisher otherwise you don't get support and you can't trust the work you are doing to be correct. And the vast majority of the cards that are up to the challenge are nvidia cards.
I have done CAD/CAM for ages, and my P3-750 with a Quadro4 700XGL isn't noticeably slower than a P4-3.4 with a Radeon X300SE running Unigraphics NX 5. I have a P3-500 with a QuadroFX-1000 card that freaking flies running CATIA V5. Again, in contrast, my 1.8GHz P4 laptop with integrated Intel graphics sucks balls running either UG or CATIA.
Speaking for the workstation users out there, please keep making high performance GPUs, Nvidia.