Questions and Answers :
GPU applications :
Nvidia just released new drivers
Message board moderation
Author | Message |
---|---|
Wee Todd Didd Send message Joined: 17 May 99 Posts: 6 Credit: 2,226,488 RAC: 3 |
Possibly people having problems can try them out and post results. http://www.nvidia.com/Download/Find.aspx?lang=en-us |
Bambi Send message Joined: 15 May 99 Posts: 26 Credit: 5,704,701 RAC: 1 |
I'll give these a go and see if CUDA behaves any better. Bambi |
Maik Send message Joined: 15 May 99 Posts: 163 Credit: 9,208,555 RAC: 0 |
Tank you for the info ... will test them. |
MeglaW Send message Joined: 21 Jun 00 Posts: 36 Credit: 479,460 RAC: 0 |
does anyone have a link to free, old version of driver cleaner? its payware nowadays but old versions would still work i guess.. downloading now, will try any second now. |
Maik Send message Joined: 15 May 99 Posts: 163 Credit: 9,208,555 RAC: 0 |
go to systemcontrol -> software -> uninstall nvidia physx & nvidia drivers -> Reboot -> windows will promt for new drivers -> cancel -> install new driver package -> Reboot i believe that should be work fine |
MeglaW Send message Joined: 21 Jun 00 Posts: 36 Credit: 479,460 RAC: 0 |
go to systemcontrol wrong, this might elev un-replaced files, the popper procedure is: 1. add remove programs and or devicemgr, uninstall old driver 2. reboot into failsafe mode, windows will promt for new drivers, say no, then run driver cleaner and remove nvidia stuff 3. reboot again into failsafe, install new driver 4. reboot normaly, and now it should work. |
Maik Send message Joined: 15 May 99 Posts: 163 Credit: 9,208,555 RAC: 0 |
MeglaW wrote: does anyone have a link to free, old version of driver cleaner? its payware nowadays but old versions would still work i guess... i found DH Driver Cleaner Professional Edition at chip.de just searched for 'driver cleaner' at google.de MeglaW wrote: the popper procedure is: Thanks, i'll consider it. |
Maik Send message Joined: 15 May 99 Posts: 163 Credit: 9,208,555 RAC: 0 |
new driver / old problem :( WU true angle range is : 0.437221 SETI@Home Informational message -9 result_overflow NOTE: The number of results detected exceeds the storage space allocated. Flopcounter: 21739917742.860546 Spike count: 30 Pulse count: 0 Triplet count: 0 Gaussian count: 0 called boinc_finish |
Byron S Goodgame Send message Joined: 16 Jan 06 Posts: 1145 Credit: 3,936,993 RAC: 0 |
new driver / old problem :( That isn't the usual AR that I get overflows in. Actually it's one of the one's I can count on not to give me those. Just curious if before you made the switch to the new driver if this was a common AR for you to have overflows in with the previous driver you were using? I noticed for myself, there's pretty much only one driver that has worked consistantly( I admit I haven't tried the new one yet), and am curious if you're finding the same? Also are the overflow still comming after a reboot? Sorry all questions and no answers in my post :) |
MeglaW Send message Joined: 21 Jun 00 Posts: 36 Credit: 479,460 RAC: 0 |
MeglaW wrote:does anyone have a link to free, old version of driver cleaner? its payware nowadays but old versions would still work i guess... perfect, altho the readme´s steps are better but along the same lines nVidia: 1. Unplug nic card or modem from net or disconnect 2. Remove nvidia from add/remove programs 3. Reboot into safe mode (F8 during boot up) 4. On reboot into safe mode WinXP SP1/SP2 will auto detect and install drivers. Just let it do its thing. If you can cancel out of the auto install hardware wizard then skip the rest and run Driver Cleaner! If you can NOT cancel out of auto detect wizard then... 5. Navigate to device manager->display adapters->highlight nVidia and click uninstall or remove but don't reboot 6. Disable any AntiVirus Software (If you don't disable is than it could give trouble) 7. Run Cab Cleaner (not needed, but recommended) 8. Run Driver Cleaner 9. Empty recycle bin and reboot 10. Install the new drivers |
Maik Send message Joined: 15 May 99 Posts: 163 Credit: 9,208,555 RAC: 0 |
Byron S Goodgame wrote: That isn't the usual AR that I get overflows in. Actually it's one of the one's I can count on not to give me those. Just curious if before you made the switch to the new driver if this was a common AR for you to have overflows in with the previous driver you were using? next one runs into a idle: WU true angle range is : 0.437221 Restarted at 17.17 percent. Flopcounter: 21501263734540.570000 Spike count: 2 Pulse count: 0 Triplet count: 0 Gaussian count: 0 called boinc_finish Want a overview on crashed wu's ? click me ;) The WU's whitch producing error's since driver update are some single one's. Next WU in row was crunched sucessfull. I didn't a restart. Can't watch my PC round the clock ... |
Byron S Goodgame Send message Joined: 16 Jan 06 Posts: 1145 Credit: 3,936,993 RAC: 0 |
I get "No access" when I click the link to the tasks you've mentioned. I'm guessing it because your computers are hidden. |
Maik Send message Joined: 15 May 99 Posts: 163 Credit: 9,208,555 RAC: 0 |
I get "No access" when I click the link to the tasks you've mentioned. I'm guessing it because your computers are hidden. damn ;) turned it on *g |
Byron S Goodgame Send message Joined: 16 Jan 06 Posts: 1145 Credit: 3,936,993 RAC: 0 |
I get "No access" when I click the link to the tasks you've mentioned. I'm guessing it because your computers are hidden. Ok, finally got a look at the tasks. First thing I noticed was that you are letting the VLAR tasks run instead of running the batch file to advert them. It's hard to say for sure, but there's a good chance they are causing your card to do the overflows in the other ARs. From what I've seen with my gpu if you did abort the VLAR's instead of running them, you'd get less false overflows in the other ARs. In my case, I've stoped all false overflows in all other ARs so far, as long as I abort the VLAR. I'm also guessing you're running the scrpt that causes stalled VLAR tasks to restart, but even if not, I think the effect the VLAR especially are having on the card , even though they give false overflows too, are still enough to still cause the negative effect on the other tasks. Anyway, thanks for letting me look at the tasks on your pc, and allowing my curiosity of those tasks. |
Maik Send message Joined: 15 May 99 Posts: 163 Credit: 9,208,555 RAC: 0 |
Ok, finally got a look at the tasks. First thing I noticed was that you are letting the VLAR tasks run instead of running the batch file to advert them. Yes i do because its not a beta project anymore. The applications was installed automaticly on my Host. There was no warnings/informations which was telling me i have to wach my PC and maybe abort tasks to avoid errors. I can live with some single 'result overflows'. |
MeglaW Send message Joined: 21 Jun 00 Posts: 36 Credit: 479,460 RAC: 0 |
it did not help stuttering windows problem. :( |
Maik Send message Joined: 15 May 99 Posts: 163 Credit: 9,208,555 RAC: 0 |
it did not help stuttering windows problem. :( CUDA causing XP to stutter every few minutes The 3. post is in english and should explain the problem you have. |
MeglaW Send message Joined: 21 Jun 00 Posts: 36 Credit: 479,460 RAC: 0 |
it did not help stuttering windows problem. :( i know, i myself have linked to that page. and i dont beleve that is the truth sence folding also uses CUDA and does not have that problem. |
Maik Send message Joined: 15 May 99 Posts: 163 Credit: 9,208,555 RAC: 0 |
depends on them how much work is send from cpu to gpu ... the feeder for gpu is running almost at 5-10% cpu-usage at my host. here is a snapshot from cuda-z while seti-cuda is runing. |
Wee Todd Didd Send message Joined: 17 May 99 Posts: 6 Credit: 2,226,488 RAC: 3 |
These new drivers seem have stopped the crashes I was getting. Mostly they were blue screens caused by one of the nvidia drivers. Completed 40 tasks without a problem. Using boincmgr 6.5.0 with enhanced AP and cuda from Raistmer. Here |
©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.