Message boards :
Number crunching :
VLAR and VHAR
Message board moderation
Author | Message |
---|---|
Tim Norton Send message Joined: 2 Jun 99 Posts: 835 Credit: 33,540,164 RAC: 0 |
Have seen a lot of discussion on the VLAR and VHAR but am curious about what the difference is over normal WU And also should you run them on CPU or GPU - is there any consensus or rules to follow? Am experimenting with a reschedule program from eFMer and you can more both of them around between the CPU and GPU Views? Tim |
Cruncher-American Send message Joined: 25 Mar 02 Posts: 1513 Credit: 370,893,186 RAC: 340 |
Have seen a lot of discussion on the VLAR and VHAR but am curious about what the difference is over normal WU In my experience, VHARs don't seem to make a difference. On the other hand, VLARs take MUCH longer on my GPUs. This is apparently due to a defect in the original implementation of the CUDA app by NVIDIA, but what exactly is wrong, I don't know. Anyone? Bueller? I shift VLARs to CPU to increase overall throughput. |
Hellsheep Send message Joined: 12 Sep 08 Posts: 428 Credit: 784,780 RAC: 0 |
What jravin said is correct. Although VHARs actually sometimes run in under 30 seconds on my GPU. I move all VLARs to CPU unless i have a lack of GPU work. Usually it takes about 20-30 minutes longer for my GPU to run VLARs and it makes the system VERY sluggish. - Jarryd |
©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.