Forum > GPU crunching
Unified installer add flops
efmer (fred):
--- Quote from: Richard Haselgrove on 07 Dec 2009, 04:40:08 am ---
--- Quote from: Claggy on 06 Dec 2009, 08:27:42 pm ---DA has added COPROCS to HOST_INFO Changeset 19797
Claggy
Edit: and after Changeset 19798 get_host_info() GUI RPC now contains GPU info
--- End quote ---
No reply to my question on boinc_alpha, but it obviously had the desired effect!
I'll grab one as soon as there's a build available, and let you see it so you can decide if it's any use.
--- End quote ---
Hopefully the info is more than the same meaningless peak flop value that's already in the message log.
Josef W. Segur:
--- Quote from: Fred M on 07 Dec 2009, 05:46:29 am ---Hopefully the info is more than the same meaningless peak flop value that's already in the message log.
--- End quote ---
The CUDA peak flops is exactly as meaningful as the older estimated flops, ratio 5.6. IMO, they are better predictors of performance than the Whetstone benchmark used for CPUs.
Joe
Raistmer:
Better prediction of relative performance I would add.
As absolute value it's meaningless of course.
9400GT performs ~ as 1 core of my quad now. And its estimation value is 45Gflops. IMO single quad core has very different actual performance...
efmer (fred):
--- Quote from: Raistmer on 07 Dec 2009, 10:28:07 am ---Better prediction of relative performance I would add.
As absolute value it's meaningless of course.
9400GT performs ~ as 1 core of my quad now. And its estimation value is 45Gflops. IMO single quad core has very different actual performance...
--- End quote ---
In the BOINC source code the comment states: FLOPS for a given app may be much less; e.g. for SETI@home it's about 0.18 of the peak
And in the flops they really use is divided by 2 or 5.
And this sort of corrections if (!strcmp(plan_class, "cuda23")) {flops *= 1.01; } for the cuda dll being more efficient.
So it maybe that newer (beta/alpha) clients, already have a correction factor for the GPU flops and the statement may be obsolete by the time of release.
Raistmer:
--- Quote from: Fred M on 07 Dec 2009, 10:45:04 am ---And this sort of corrections if (!strcmp(plan_class, "cuda23")) {flops *= 1.01; } for the cuda dll being more efficient.
--- End quote ---
CUDA 2.3 bring better performance improvement for CUDA MB than 1% so those 2 plans definitely will have different peak performance to real performance ratios for SETI project.
And 45/5=9(Gflops) - too much for single quad core performance too IMO.
That is, 5 times "peak to real" is pretty arbitrary coefficient.
All htis makes absolute Gflops unusable IMO. What needs to be checked: will 2 cards rated 45 and 90 Gflops by BOINC (for example) exhibit 2 times difference in performance for CUDA MB app or not (can BOINC estimates be used for relative performance comparisons or not).
Navigation
[0] Message Index
[#] Next page
[*] Previous page
Go to full version