Questions and Answers :
GPU applications :
While I have plenty of cpu tasks, I have no Gpu tasks and its coming up on maintence tuesday?
Message board moderation
Author | Message |
---|---|
Tom M Send message Joined: 28 Nov 02 Posts: 5124 Credit: 276,046,078 RAC: 462 |
Seems to be a little odd that I have no Seti GPU tasks. I do have 2 GPUGRID tasks (which run 12 hours or more a piece) One is in process and is down below 10 hours left. So is there a reason, besides none being available, that I am not getting any seti gpu tasks? What am I missing? Thanks, Tom A proud member of the OFA (Old Farts Association). |
OzzFan Send message Joined: 9 Apr 02 Posts: 15691 Credit: 84,761,841 RAC: 28 |
Resource scheduling and the debt system would be a good reason why. What are your resource shares for other projects and how do they balance out percentage-wise? |
Tom M Send message Joined: 28 Nov 02 Posts: 5124 Credit: 276,046,078 RAC: 462 |
Resource scheduling and the debt system would be a good reason why. What are your resource shares for other projects and how do they balance out percentage-wise? I was running Seti at 100(%?) cpu/gpu. SetiBeta at 1% cpu only, WorldGrid at 1% cpu only apparently, GPUGrid at 1% long GPU tasks (12+ hours) and Cosmology at 1% cpu only apparently. I had 2 long Gpu tasks from Gpugrid. After looking at some other threads I have switched everyone but Seti to 0%. This should cause GPUGrid to only download when I run out of GPU tasks. SetiBeta at 0, now is cpu/gpu. So I am guessing that maybe Gpugrid caused Seti to not send any gpu tasks? Tom A proud member of the OFA (Old Farts Association). |
Jord Send message Joined: 9 Jun 99 Posts: 15184 Credit: 4,362,181 RAC: 3 |
The resource share is a long term value, it will always take a while for the value changes to propagate. But the RS also shows which project to run next and for how long. When you had Seti set to 100 and the other projects to 1, (not percentages as there can only be 100% and not more), that meant that eventually the other projects would want to use the CPU for about as long as you had allowed Seti to run work. You now set the values to 0%, which will mean that any work BOINC still has in cache for the projects at 0% will have to be done first, because they are now set to be 'emergency projects' only, where work is gotten from when the main project has no work. With the estimated run time of the two GPUGrid tasks being over a day, it's no use for BOINC to try to get more work at this moment, as it's not even sure it can finish the tasks in cache by their individual deadline. And there's no use getting more work in, when the present work keeps the hardware occupied. |
©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.