| Impulsivity said: It might be good, it really depends on how it is used and how much it costs. They compare it to a last generation 285 (the 385 should be out shortly with much more competitive numbers much like the 5870 blew away all the last gen processors when it came out) and it does a bit better, yes, but not really that much better. I expect that the Larrabee will be matched roughly by the high end of Nvidia's next gen GPUs. Additionally they are comparing a theoretical pull out all the stops prototype with an actual consumer product that has price constraints (in the case of the 285 around 300 dollars). I'd like to see how a production 300 dollar Larrabee model does on benchmarks, I bet it would be nowhere near as high. |
here is the big point:
unlike theoretical numbers that are usually disclosed by ATI and nVidia - this was an actual SGEMM benchmark calculation used in the HPC community.
even if it was a prototype
The keynote continued while the engineers scrambled at the back to try to beat the 1TFLOPS barrier. A couple of minutes before the end of the keynote, Justin added the infamous "And one more thing…" Initial overclocked performance was 913 GFLOPS, moved slowly past 919 GLOPS, bounced up to 997 GFLOPS and ultimately passed the 1TFLOPS barrier with 1006 GFLOPS. Now, we can debate the numbers all we want, but the fact of the matter is that nVidia Tesla C1060 delivers only 370 GFLOPS in an identical SGEMM 4Kx4K calculation. Thus, Larrabee today comes at 2.7x math performance of GT200 chip.
the test showed is overclocked, the fact that the Chip could handle that in overclock. now what the stable and direct benchmark will be once the Chip get's released would be neat to see what it will benchmark at is down the road.
and what look's to be pretty darn impressive result's even if they are reduced

I AM BOLO
100% lover "nothing else matter's" after that...
ps:
Proud psOne/2/3/p owner. I survived Aplcalyps3 and all I got was this lousy Signature.







