Hardware IDF Day 1 - Sean Maloney demos Larrabee

Larabees a bit of a disspointment in terms of its ability solidly chuck out the fps, i was expecting more, especially with some of the hype on the internet, and intels own bragging.

But at least its a step forward for ray tracing, if not a pointless one for the consumer market

 

Larrabee is not really about producing a graphics card. Its about producing a highly parallel set of CPUs with good floating point throughput that is x86 compatible. The graphics card vendors have seen fit to start to target the high performance market, traditionally Intel's market and this is Intel's response to CUDA, OpenCL and DirectCompute running on ATI/NVidia graphics cards. I don't think the end goal is to run a full directX game on this architecture yet, just to secure the HC market against the intrusions of the less capable graphics cards. This is a market that Intel wants to dominate as its only now emerging.

One such program running on that architecture is a graphics rendering engine and ray tracing is a much simpler engine to implement to make it look good than a fully raster capable one. I expect we'll see other technical demos from Intel as we approach the launch that are all about high computation and massively parallel problems, with a focus on an API much easier to use than OpenCL and ilk.

Ray tracing and graphics rendering is in many ways a warning to the graphics card vendors, especially NVidia. The war of words 2 years ago was that NVidia was going to eat Intel with its massively parallel stream processors, now Intel is not only producing a significantly faster and more capable multi core system but its also showing how one day it'll be able to eat NVidia's market.

So in a nutshell don't expect Larrabee to really beat out the 5870, its likely to be an order of magnitude slower in game rendering. However in the high compute market Larrabee looks like its going to dominate. Moore's law will then (Intel hopes) take it to eating the NVidia/ATI graphics market with engines that are fully programmable and with higher fidelity with Ray Tracing, all implemented in Intel's key IP - x86.

 

Every piece of Intel material would disagree with that statement. The first products are "discrete graphics" and support DirectX, OpenGL and OpenCL (as well as the native programming mode) which, if I'm not mistaken, means it's a graphics card. I wish people would stop trying to say it's something else.

 

Time is on Intel's side, that's the reason for "Intel's obsession with ray tracing". As cores and threads increase with each new generation of CPUs, ray tracing will approach viability and, eventually, even surpass the DirectX version of the day. It's a problem to be solved by adding muscle. Yes, right now it's nothing but marketing hype: Larrabee won't even come close to compete with what ATI and Nvidia had last year, much less with future GPUs. But Intel have deep pockets and Larrabee is but the first step. A very important first step.

 

I think tim has a point that kind of stuck with me least.. ray tracing is wasteful- might be easier on the programmer but it takes alot more power to render the same scene

 

I'm not sure what Intel is trying to do with Larrabee. It almost looks like Intel doesn't know what Intel is doing with it either.

Raytracing is amazingly difficult and computationally expensive to do right, but it seems like the demo was solely to 'justify' Larrabee existance at the moment. I agree that the pseudoGPU actually has relevance, but where it will fit in eventually is a bit of a crap shoot.

At 32 cores, Larrabee hits about 2 SP TFLOPs @ 2GHz whereas the Radeon 5870 sits in at 2.72TFLOPs. If Intel is going down the "graphics card" road with this, they'll have one heluva problem...mainly, their GPU doesn't compete at the moment.

Of course, to hit the same theoretical bandwidth, you're looking at a 48-core Larrabee at 2GHz* using quite a bit more power (assuming both hit maximum theoretical performance on both cards and being generous on the power specs for Larrabee).

* - Assuming that Larrabee scales at 100% to 32 cores, and 90% after that.

 

I expected it. Larabee always looked like a flop in the making. I wont go into the reasoning, but this news is no surprise for me.

 

There doesnt seem to be anything faster than Nvidias products on display here...

Will it? Doesn't the DX11 standard blur the gpgpu part of the equation? Neither ATI and especially not Nvidia are lagards at this kind of thing, and from what i can see Rv870 and GT300 are going to really shake up things up regarding gpgpu functionality, at least in the desktop environment. Theyre taking gpu programability a leap forward. Come Larabees launch video transcoding and other such gumpf one can do at home on their dx 11 gpu will already be super quick with super quality. I fail to see at least in the home context where Larabee will 'fit in'.

Larabee looks to be destined for the hpc crowd at the rate its going, and not be worth the wait neither.

 

Na. Reflective surfaces dont pull in the gamers that much. Once its viable to ray trace games, imagine what will be possible with rasterization. As in take a step back 5 years in quality for your ray traced scene, or have individually animated ants in ant hills in Crysis 5.

(disclaimer: ants not best example.)

 

I think a combination of Raster and Raytraced is the way to go. I’m excited for this prospect in the new DirectX. I just hope DX11 really takes off and we start getting back to having some stunning looking PC games.
As good as the consoles are for the rate in which new games are coming out their dated technology is really holding back the PC gaming market. There are only a handful of PC only games out nowadays compared to the old days. Its sad. I will be looking forward to the new consoles solely to see if they support DX11 so PC can progress again. What current games (consol ports) can't be easily maxed out in 1080p by a GTX280 (other than Crysis)? Until we have the software/games in development to support all these new ideas & GPU's they're all pretty pointless IMO. The biggest issue here is to get the PC market selling well again. If you pirate games don’t! Your killing the industry/hobby you like just because you won’t spend £25.. Crazy

 

Its not. Its very likely a "mainstream" or "mainstream-performance" card. Not the top end "enthusiast" card.

While Haswell is for desktops and notebooks...Not sure if they're gonna put Larrabee technology into the netbook market.

Pretty much. The overall battle is about processing power. It doesn't matter where you get it: CPU, GPU, etc. And yeah, Intel does want it all to be x86 based.

It is marketed as a graphics card...Read between the lines.

Strictly speaking, its a flexible solution for both GPU and GPGPU roles. Its really a bunch of custom x86 cores with a software rendering stack that is capable of doing 3D via rasterization and ray-tracing. Its not like a traditional GPU (where some bits are hardcoded)...Its getting x86 architecture to do other things. Their developer/programming information pretty much conveys this general message.

Its just that we don't know their real intention in any detail. That's all. They've just given us a general direction. (Which is somewhat vague).

Think of heterogeneous processors like "Haswell" in 2012.

Its not going to be gunning for the enthusiast market like the Radeon HD 5870 is.

Think mainstream/mainstream-performance markets. That's where money is made when selling graphics cards.

Rumours indicate 32 cores at 45nm process, while a future version (Larrabee 2) will be 48 cores at 32nm process.