Message boards :
Number crunching :
Long GPU units again?
Message board moderation
Previous · 1 · 2 · 3
Author | Message |
---|---|
Helli_retiered Send message Joined: 15 Dec 99 Posts: 707 Credit: 108,785,585 RAC: 0 |
Nvidias don´t like VLAR´s. Another funny thing is - if you cancel these wrong assigned Workunits you get them back as Resend lost Workunits. But now on the CPU. :-) Helli A loooong time ago: First Credits after SETI@home Restart |
Mike Send message Joined: 17 Feb 01 Posts: 34258 Credit: 79,922,639 RAC: 80 |
Nvidias don´t like VLAR´s. True, noticed that also a couple weeks ago. With each crime and every kindness we birth our future. |
bloodrain Send message Joined: 8 Dec 08 Posts: 231 Credit: 28,112,547 RAC: 1 |
well i been not getting any long wu with a gts250 |
Fred J. Verster Send message Joined: 21 Apr 04 Posts: 3252 Credit: 31,903,643 RAC: 0 |
After a lot of problems with my ATI host,(HD4850, also burned) and HD5000 series ATI, would not work. At least not with MB WU's. And both (ASUS P5E)Mobo's will not work correct, with 2 nVIDIA cards, I've taken out the HD5870 and replaced it with the GTX470, from my XP64 rig. Reïnstalled BOINC and LUNATICs V0.37 (FERMI Preview),so I can crunch the MB tasks on this host, before they expire. First thing I noticed, a VLAR WU, cunched by the FERMI (GTX470), now 1 at a time. Cause I wanna know, if it's capable of crunching one, in 'normal' way and times, estimated at 12 hours! (Which is normal, cause BOINC Mngr, doesn't 'know' this GTX470 card. When BOINC started up, did his Bench Mark Check and started crunching. I now see, it is at 30% after 17 minutes and happy crunching along. If it finishes, within reasonable time, or even less time then the CPU, I'll let them finish this one and maybe also other vlar WU's. Have to find an other mobo, for my FERMI's(GTX470 & 480), cause these 2 (P5E-ASUS) mobo's, don't work well with nVIDIA GPUs, also uses PCI-E x 16 ver.1.0 (ATI card(s) are well accepted and even 2 will run in PCI-E x 16 ver.2.0!) The one vlar, it's crunching now on the 470GPU, is atleast going 4x faster, compaired to a similar, 0.01AR MB task. I would dare to try this on my GTS250, but FERMIs, are quite different. This WU, could not be 'matched', but looks OK?! Device 1: GeForce GTX 480 is okay SETI@home using CUDA accelerated device GeForce GTX 480 Priority of process raised successfully Priority of worker thread raised successfully size 8 fft, is a freaky powerspectrum size 16 fft, is a cufft plan size 32 fft, is a cufft plan size 64 fft, is a cufft plan size 128 fft, is a cufft plan size 256 fft, is a freaky powerspectrum size 512 fft, is a freaky powerspectrum size 1024 fft, is a freaky powerspectrum size 2048 fft, is a cufft plan size 4096 fft, is a cufft plan size 8192 fft, is a cufft plan size 16384 fft, is a cufft plan size 32768 fft, is a cufft plan size 65536 fft, is a cufft plan size 131072 fft, is a cufft plan ) _ _ _)_ o _ _ (__ (_( ) ) (_( (_ ( (_ ( not bad for a human... _) Multibeam x32f Preview, Cuda 3.0 Work Unit Info: ............... WU true angle range is : 8.092281 Flopcounter: 10028718023014.631000 Spike count: 9 Pulse count: 0 Triplet count: 8 Gaussian count: 0 called boinc_finish </stderr_txt> ]]> It appears, the VLAR MB task is crunched, reported and uploaded, so I wasn't able to see the time, it took, about an hour. It is in my SETI BÊta account, have to track it down and see the result. |
-BeNt- Send message Joined: 17 Oct 99 Posts: 1234 Credit: 10,116,112 RAC: 0 |
Fred I've been trying the same experiment today and watching the times on vlar's on my GTX480. It crunched 1 vlar in 1 hour 2 minutes and is crunching 2 now and it looks to get two done in 1 hour 13 minutes. However it makes my PC lag a good bit crunching two of those at a time. Normal completion times for a vlar on my cpu is about 2 hours 30 minutes to 2 hours 40 minutes on comparable units. Pretty cool. Traveling through space at ~67,000mph! |
©2024 University of California
SETI@home and Astropulse are funded by grants from the National Science Foundation, NASA, and donations from SETI@home volunteers. AstroPulse is funded in part by the NSF through grant AST-0307956.