Nvidia Claims Intel's Larrabee Is "a GPU From 2006"

Barence sends this excerpt from PC Pro: "Nvidia has delivered a scathing criticism of Intel's Larrabee, dismissing the multi-core CPU/GPU as wishful thinking — while admitting it needs to catch up with AMD's current Radeon graphics cards. 'Intel is not a stupid company,' conceded John Mottram, chief architect for the company's GT200 core. 'They've put a lot of people behind this, so clearly they believe it's viable. But the products on our roadmap are competitive to this thing as they've painted it. And the reality is going to fall short of the optimistic way they've painted it. As [blogger and CPU architect] Peter Glaskowsky said, the "large" Larrabee in 2010 will have roughly the same performance as a 2006 GPU from Nvidia or ATI.' Speaking ahead of the opening of the annual NVISION expo on Monday, he also admitted Nvidia 'underestimated ATI with respect to their product.'"

Doh of the Day

...he also admitted Nvidia 'underestimated ATI with respect to their product.'

Good, learn from that and don't make that same mistake again!

Larrabee [...] will have roughly the same performance as a 2006 GPU from Nvidia or ATI.'

DOH!

Re:Doh of the Day

Intel has made some bad mis-steps in the past, and one of them was failing to design their processors around the strengths and weakness' of their memory architecture. Rambus is a prime example. It was a superior solution for the wrong problem, and Intel failed to design their processors to take advantage of the memory's strengths, and it looks like they are doing it again. The limiting factor in CPU / GPU performance isn't how many instructions you can pound into any given second, its how much total memory can you get at, in that time frame. It does you no good to be able to process 16 billion pixels / second, when you can only get the data for 4 billion per second from your memories. Better to build a system that can get 6 Billion per second from the memory, and can process only 6 billion per second. That is the fundamental problem that Nvidia seems to understand, and Intel doesn't.

-=Geoskd

Repudiating my own quote

Although I appreciate the attention from NVIDIA and Slashdot, I can't support that alleged quote from my blog (http://speedsnfeeds.com).

First, what's being described as a quote is actually just John Montrym's summary from my original post, which is here:

http://news.cnet.com/8301-13512_3-10006184-23.html [cnet.com]

What I actually described as equating to "the performance of a 2006-vintage... graphics chip" was a performance standard defined by Intel itself-- running the game F.E.A.R. at 60 fps in 1,600 x 1,200-pixel resolution with four-sample antialiasing.

Intel used this figure for some comparisons of rendering performance. If Larrabee ran at 1 GHz, for example, Intel's figures show that it would take somewhere from 7 to 25 Larrabee cores to reach that 60 Hz frame rate.

Larrabee will probably run much faster than that, at least on desktop variants.

Well... rather than writing the whole response here, I think I'd rather write it up for my blog and publish it there. Please surf on over and check it out:

http://news.cnet.com/8301-13512_3-10024280-23.html [cnet.com]

Comments are welcome here or there.

. png

no wonder its slow

No wonder it's so slow. He keeps making reference to how it paints things. Can't move on to another frame until the previous one has dried.

Re:no wonder its slow

Maybe the imp died

Intel isn't aiming at gamers

So why is NVIDIA on the defensive?

Intel is aiming at number crunchers (note that their chip uses doubles, not floats). They don't want NVIDIA to steal that market with CUDA.

When Intel says "graphics", they mean movie studios, etc.

If Larrabee eventually turns into a competitor for NVIDIA, all well and good, but that's not their goal at the moment.

Re:Intel isn't aiming at gamers

Ok. Nvidia and AMD/ATI probably will overperform the Intel GPU. However Intel has open specs of their GPU-s, so for NOT the gamers there will be a quite good performance GPU to be used under Linux for various purposes with high quality OSS drivers. Im looking forward to it.
Shinatosh

Re:Intel isn't aiming at gamers

So why is NVIDIA on the defensive?

I think the Nvidia people think pretty highly of themselves, rightfully so, and Intel has recently been making a number of bold claims, without backing them up. In a poker analogy, Nvidia is calling Intels bluff.

Re:Intel isn't aiming at gamers

It's obvious that the graphics angle is really just a Trojan horse; they're using graphics as the reason to get it into the largest number of hands possible, but what they really want to do is to keep people writing for the X86 instruction set, rather than OpenCL, DirectX 11, or CUDA. Lock-in with the X86 instruction set has served them too well in the past.

In other words, general compute was an area where things were slipping out of their grasp; this is a means to shore things up.

It's a sound business strategy. But I have to agree with blogger-dude; I don't see them being overly competitive with NVIDIA and AMD's latest parts for rasterized graphics anytime soon.

Re:Intel isn't aiming at gamers

No, Intel has been very clear that it is targeting games, even saying "we will win" about them, see this interview with an Intel VP [arstechnica.com].

NVidia is on the defensive for the simple reason that it needs to be. Not because Intel has a product that threatens NVidia, but because Intel is using classic vaporware strategies to undermine NVidia (and AMD/ATI). Intel is basically throwing around promises, and by virtue of its reputation a lot of people listen and believe those promises. With 'amazing Intel GPU technology just around the corner', some people might delay buying NVidia hardware. NVidia is trying to prevent that from happening.

Re:Intel isn't aiming at gamers

Uh, nVidia has supported Linux, FreeBSD and Solaris for some years now. The drivers are binary-only, which makes them unacceptable to some people (I'd prefer not to run them, personally), but if you're buying a Dell with Ubuntu or a Sun with Solaris on it you can easily use an nVidia GPU and get the same sort of performance you would from Windows.

Re:Intel isn't aiming at gamers

zquote>Right now they're crapping their pants because they gambled everything on Vista which is failing spectacularly, whereas both ATi and Intel have got a 2 year head start on supporting Ubuntu out of the box.

Traditionally (before the AMD buyout) ATI has had terrible support under Linux, nVidia has been delivering their binary drivers and they have in my experience been easy to install, stable and fully functional much longer than ATI. By the way, the latest 177.68 drivers should now have fixed the recent KDE4 performance issues. From what I've understood the fglrx (closed source) ATI driver has made considerable progress so maybe now they're equal, but closed vs closed source nVidia got nothing to be ashamed of. For exampel, ATI just this month added CrossFire support while SLI has been supported since 2005, that's more like three years behind than two years ahead.

Of course, ATI is now opening up their specs but it's going slowly. For example, they have not yet recieved [phoronix.com] the 3D specs on the last generation R600 cards, much less the current generation. And after those specs are released some very non-trivial drivers must be written as well, meaning it could take another few years before we see fully functional open source drivers. Also this strategy is less than a year old, so if they're two years ahead they're moving fast. Nothing of this is something that should make nVidia executives the least bit queasy.

They are crapping their pants because ATI has delivered two kick-ass parts in the 4850 and 4870, and there's very little reason to buy anything else these days. They are crapping their pants because Intel won't give them a Nehalem system bus license. They're crapping their pants because the 800lb gorilla that's Intel is entering a market everyone else has been leaving. They're crapping their pants because the US economy is shaky and people might not spend big $$$ on graphics cards. But over Linux support? Put it into perspective, and you'll see it's an extremely minor issue.

Re:Intel isn't aiming at gamers

Both Intel and nVidia - proprietary driver companies - should be on defensive right now.
You obviously don't know much about Intel's commitment to open-source in its drivers. Instead of recently dumping some of the specs to its chips to the community at large, Intel has actively paid developers to maintain top-quality 100% FOSS drivers for Linux and X11 for years, making its commitment light-years ahead of ATI or Nvidia. Ever hear of Keith Packard, you know, the leading developer of the entire X system? He's an Intel employee. For all the accolades that ATI gets for dumping a bunch of specs on the web, Intel has put vastly more time & money into supporting OSS, but still gets labeled as "closed source" by fanboys.

Gee, How "Forward Thinking" of You, NVidia!

"OH MY GOD! CPU AND GPU ON ONE DIE IS STOOOOOOOOPIIIIIDDDDDEDEDDDD!!!1111oneoneone"

How stupid is it really? So what if the average consumer actually knows very little about their PC. That doesn't necessarily mean it won't be put into a person's PC.

If they were really forward thinking, they could see it as an effort to bridge the gap between low-end PC's and high-end PC's. Now maybe, at some point in the future, people can do gaming a little better on those PC's.

Instead of games being nigh unplayable, are now running slightly more smoothly. With advance in this design, it could really work out better.

Sure, for the time being, I don't doubt that the obvious choice would be to have a discrete component solution for gaming. However, there might be a point where that isn't in the gamers best interests anymore. I'm not a soothsayer, I don't know.

Still, I can't only help but imagine how Intel's and AMD's ideas can only help everyone as a whole.

Larrabee as a rasteriser...

A recent journal article on ArsTechnica points to an Intel blog on Larrabee: http://arstechnica.com/journals/hardware.ars/2008/05/01/larrabee-engineer-on-personal-blog-larrabee-is-all-about-rasterization [arstechnica.com] Curious.

Classic case of disruption

Ten years ago you would see Nvidia GPUs in everything from low- to high-end. Today, not so much - Intel dominates the low-end spectrum, with ATI hanging onto a somewhat insignificant market share. The Larrabee is Intel moving upmarket. Sure, it might not perform as well the latest Nvidia or ATI high-end GPU but it might be enough in terms of performance or have other benefits (better OSS support) to win some of Nvidia's current market share over. Considering it's supposedly the Pentium architecture recycled, it's also reasonable to assume the design will be relatively cost-effective and allow Intel to sell at very competitive prices while still maintaining healthy profit margins.

It's a classic case of disruption. Intel enters and Nvidia is happy to leave because there's a segment above that's much more attractive to pursue. Continue along the same lines until there's nowhere for Nvidia to run, at which point the game ends - circle of disruption complete. See also Silicon Graphics, Nvidia's predecessor in many ways.

Re:Classic case of disruption

>ATI hanging onto a somewhat insignificant market share.

C'mon, 17 million units shipped in a quarter and ~20% of the market is hardly 'a somewhat insignificant market share' in a market with four major players (Intel, nVidia, VIA).

For comparison, take Matrox, they have insignificant market share with about 100K/q

Re:Classic case of disruption

Ten years ago, the Riva TNT was yet a few months away. S3 and ATI both had a great marketshare for low to mid-end, and 3dfx dominated the very top segment for gamers.

AMD is in the Best Position

Lots of people here and analysts have written off AMD. I think AMD is in a great position if they can survive their short term debt problems which is looking increasingly likely.

Consider the following:

• Intel's GPU tech is terrible.
• Nvidia doesn't have an x86 design / manufacturing experience, x86 license, or even x86 technology they want to announce.
• AMD currently has the best GPU technology and their technology is very close to Intel's for CPUs.

AMD is in a great position like no other company to capitalize on the coming CPU / GPU convergence. Everyone jeered when AMD bought ATI but it is looking to be a great strategic move if they can execute on their strategy.

AMD has the best mix of technology, they just have to put it to good use.

Re:AMD is in the Best Position

Well, if you read the reviews, AMDs integrated graphics sollution 780g kicks ass. Only the very very newest Intel integrated chipset is slightly better, but that uses around 20W compared to AMD's chipset's 1W

Re:AMD is in the Best Position

Nvidia doesn't have an x86 design / manufacturing experience, x86 license, or even x86 technology they want to announce

True. They do, however, have an ARM Cortex A8 system on chip, sporting up to four Cortex cores and an nVidia GPU in a package that consumes under 1W (down to under 250mW for the low-end parts). Considering the fact that the ARM market is currently an order of magnitude bigger than the x86 market and growing around two-three times faster, I'd say they're in a pretty good position.

Re:AMD is in the Best Position

Nvidia does indeed have license to x86. They acquired it when they bought all of 3dfx's intellectual property. They in fact manufacture a 386SX clone. Rumors have been persisting that they are looking to enter the x86 market. It should be noted that they are still relative outsiders in that their licensing doesn't extend into the x86-64 instruction set, which is taking over the market now.

What bullshit.

From the SIGGRAPH paper they need something like 25 cores to run GoW at 60Hz. That's 1Ghz cores for comparison though. LRB will probably run at something like 3Ghz, meaning you only need like 8-9 cores to run GoW at 60, and with benchmarks stretching up to 48 cores you can see that this has the potential of being very fast indeed.

More importantly, the LRB has much better utilization since there aren't any fixed function divisions in the hardware. E.g. most of the time you're not using the blend units. So why have all that hardware for doing floating point maths in the blending units when 99% of the time you're not actually using it? On LRB everything is utilized all the time. Blending, interpolation, stencil/alpha testing etc. is all done using the same functionality, meaning that when you turn something off (like blending) you get better performance rather than just leaving parts of your chip idle.

I'd also like to point out that having a software pipeline means faster iteration, meaning that they have a huge opportunity to simply out-optimize nvidida and amd, even for the D3D/OGL pipelines.

Furthermore, imagine intel suppyling half a dozen "profiles" for their pipeline where they optimize for various scenarios (e.g. deferred rendering, shadow volume heavy rendering, etc. etc.). The user can then try each with their games and run each game with a slightly different profile. More importantly, however, is that new games could just spend 30 minutes figuring out which profile suits them best, set a flag in the registry somewhere, and automatically get a big boost on LRB cards. That's a tiny amount of work to get LRB-specific performance wins.

The next step in LRB-specific optimizations is to allow developers to essentially set up a LRB-config file for their title with lots of variables and tuning (remember that LRB uses a JIT compiled inner-loop that combines the setup, tests, pixel shader etc.). This would again be a very simple thing to do (and intel would probably do it for you if your title is high profile enough), and could potentially give you a massive win.

And then of course the next step after that is LRB-specific code. I.e. you write stuff outside D3D/OGL to leverage the LRB specifically. This probably won't happen for many games, but you only need to convince Tim Sweeney and Carmack to do it, and then most of the high profile games will benefit automatically (through licensing). My guess is that you don't need to do much convincing. I'm a graphcis programmer myself and I'm gagging to get my hands on one of these chips! If/when we do I'll be at work on weekends and holidays coding up cool tech for it. I'd be surprised if Sweeney/Carmack aren't the same.

I think LRB can be plenty competitive with nvidia and amd using the standard pipelines, and there's a very appealing low-fricion path for developers to take to leverage the LRB specifically with varying degrees of effort.

I'f prefer stable releae from 2006...

With all OpenGL extensions supported working properly, to latest and greatest from NVIDIA where I can never be sure which extension work on which driver with which card.

Re:Better than NVIDIA's proprietary hardware

Sorry, but I did exactly that, and got bitten recently: NVidia's drivers for old graphics cards lag behind more and more. I can no longer update one of my systems because the ABI version for their GLX doesn't get updated.
The fix would be trivial (just recompile the current version), but Nvidia clearly would rather sell me a new card.