GLeeM
Elite Member
- Apr 2, 2004
- 7,199
- 128
- 106
Oops! I hope they aren't any of mineThe old units I've received have all be re-sends: errors and aborts.
I looked at some of mine and my wingmen crunched them with SSE3 app. I don't get it.
Oops! I hope they aren't any of mineThe old units I've received have all be re-sends: errors and aborts.
hmm...well now i'm confused. the task that ran for 8,951s was a v100.00 task. i've since received a mix of SSE3-optimized and non-SSE3-optimized tasks, but the non-SSE3-optimized tasks are now v101.10, and only take ~4,550s (~1.25 hours) to complete...so the SSE-optimized tasks are only twice as fast as these v101.10 non-SSE3 tasks, but three times as fast as the v100.00 non-SSE3 tasks. i'm not sure what the difference between the two non-SSE3 tasks are...i ran a single Asteroids@Home task on my 1055T approx. one month ago (pre-SSE3 optimization), and it completed in 8,951s (~2.5 hours). ran some SSE3-optimized tasks today on the same CPU and they finished in ~2,800s (~47 minutes). its unfortunate that the 1055T won't see even more improvement when the AVX-optimized app releases, but at the same time i can't complain about my run times being only a third what they used to be on an AMD Phenom II CPU.
hmm...well now i'm confused. the task that ran for 8,951s was a v100.00 task. i've since received a mix of SSE3-optimized and non-SSE3-optimized tasks, but the non-SSE3-optimized tasks are now v101.10, and only take ~4,550s (~1.25 hours) to complete...so the SSE-optimized tasks are only twice as fast as these v101.10 non-SSE3 tasks, but three times as fast as the v100.00 non-SSE3 tasks. i'm not sure what the difference between the two non-SSE3 tasks are...
Good news!as preparation step for nVidia CUDA development.
Posted in this thread at asteroids message board: http://asteroidsathome.net/boinc/forum_thread.php?id=169#1546
"No, mix of SSEx and AVX instruction is very ineffective. Especially when SSE2 instruction follow AVX instruction.
We will not release AVX version, simply because it's slower than SSE3. The final approach will be:
1. Standard app
2. Pure SSE2 app
3. SSE3 app (the fastest one)
Kyong is testing SSE2 now. I'm working on standard app now (some backports from sse3 version) as preparation step for nVidia CUDA development."
I have inspect this and it looks like AVX2 app for Intel Hasvel.
AVX2 brings new integer instructions to 256 bit AVX world which is missing in AVX. We use them in app so our AVX app must use SSE2 instructions for integers.
We use Visual studio 2010 for win builds and there is no AVX2 support. Visual studio 2012 do not support Win Vista and older OS.
I have ordered one i5-4670 in our company and I will test AVX2. If tests will be succesfull we will create download section and let users download special app with app_info.xml included.
I will do CUDA version for nVidia cards so I can talk about CUDA version only.Only CUDA, no OpenCL?
Kyong has some long term plans about OpenCL, but I can't say more.
Yey got WUs now , just gotta let F@H finish it's current WU, currently @85%.
Anyone know how to stop it getting more WUs?
Dusted out & now down to @ 66C, 100% load, ambient 25C.
Hmm, CPU temp still quite high though.
Anyone else notice higher temps with A@H?
Is anyone else still struggling to get WUs?
I'm gonna have to ramp down to 50%.
This reminds of the old days of SETI lol.