I have a 1gb nvidia geforce 210, overclocked to 620 mhz and the shader overclocked to 1436 mhz, it has 24 shaders, CUDA.

My problem is I'm not given enough crunch time to run gpugrid! A "long run", which is all i get which is 8-12 hours on the "fastest cards" takes my card 194 hours straight OVERCLOCKED! and I'm only given 5 days to complete it. This is with the system on 24/7, the unit running high priority and as stated with the GPU overclocked to max capacity.

Is there a way to not receive the "long runs" or am I screwed here? I don't have $400-$4000 to buy a super fast video card but I want to help. Do I have any options or should I just kill GPUGRID on my BOINC client?

Hi Chris,

You can select to run only the "ACEMD standard" WUs (not long) in your GPUGrid Preferences

http://www.gpugrid.net/prefs.php?subset=project

Thanks,
Steve

---

Thanks - Steve

Thanks, I don't know how i missed that in the prefs :)

I'm sorry to inform you, but your GPU is too slow even for the short tasks. It won't finish a short task in 3 days, after this "pre-deadline" the server will reissue your workunits to a faster host. So, your GPU will genereate many-many reissue, so you are basically just wasting electricity with GPUGrid.