Message boards :
Graphics cards (GPUs) :
195.62 drivers
Message board moderation
| Author | Message |
|---|---|
|
Send message Joined: 24 Dec 08 Posts: 738 Credit: 200,909,904 RAC: 0 Level ![]() Scientific publications ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
I'm testing the 195.62 WHQL drivers at the moment on an XP machine with a GTX295. For those that don't know the 195.62 drivers support CUDA 3.0 and OpenCL 1.0 code. I am currently using the cuda 2.3 DLL's. The cuda 3.0 DLL's are only available to registered nvidia developers at the moment. So far i've run some seti cuda work without any issues (apart from a couple of VLAR's that I aborted). I have now picked up 2 GPUgrid tasks and they have started, so i'll know in about 6 hours how it goes. BOINC blog |
|
Send message Joined: 6 Jun 08 Posts: 152 Credit: 328,250,382 RAC: 0 Level ![]() Scientific publications ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]()
|
I have the same GTX-card and driver more than a week and all works fine. Seti - Collatz - Milky Way - Einstein - GPU grid. All OK. I also have one machine with GTX 260. I cannot install this driver 195.62 so i have to stay on 191.07. Any reason already known? |
|
Send message Joined: 21 Dec 08 Posts: 51 Credit: 26,320,167 RAC: 0 Level ![]() Scientific publications ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]()
|
2, EVGA 260s here. Win. XP64, Nvidia 195.62 runs good. Although I have not had any problems with my 260s ever, with any driver. Except overclocking too much. They are 55nm 260s. |
|
Send message Joined: 24 Dec 08 Posts: 738 Credit: 200,909,904 RAC: 0 Level ![]() Scientific publications ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
I have now picked up 2 GPUgrid tasks and they have started, so i'll know in about 6 hours how it goes. Well they completed just fine. I'll upgrade the Win 7 (64 bit) machine and see how it goes next. Maybe the project can get a cuda30 plan class organised so we can use cuda 3.0? BOINC blog |
|
Send message Joined: 24 Dec 08 Posts: 738 Credit: 200,909,904 RAC: 0 Level ![]() Scientific publications ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
I'll upgrade the Win 7 (64 bit) machine and see how it goes next. And that worked as well. Next i'll try my dual GTX260 machine. They haven't worked with GPUgrid for some time, so i'm not expecting it to work but you never know. According to GPU-Z they have 65nm G200 chips. BOINC blog |
|
Send message Joined: 24 Dec 08 Posts: 738 Credit: 200,909,904 RAC: 0 Level ![]() Scientific publications ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
Next i'll try my dual GTX260 machine. They haven't worked with GPUgrid for some time, so i'm not expecting it to work but you never know. According to GPU-Z they have 65nm G200 chips. As expected that didn't work. Got the usual cufft errors. So I guess they'll stick with other projects for the time being. BOINC blog |
©2025 Universitat Pompeu Fabra