| Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
| World Community Grid Forums
|
| No member browsing this thread |
|
Thread Status: Active Total posts in this thread: 26
|
|
| Author |
|
|
Bearcat
Master Cruncher USA Joined: Jan 6, 2007 Post Count: 2803 Status: Offline Project Badges:
|
REBOOT THE PC, in Windows, I am assuming you are using Windows, has NO WAY to reset the gpu without a full restart of the pc. Linux CAN reset the gpu but Windows can NOT!! This often fixed the problems but if not check for gpu driver updates and IF you have very recently done an update and are at the most recent try rolling back to an older version. Gpu drivers re first and foremost for gamers, and that is NOT always beneficial to us crunchers. According to this document, Win7 will reset the GPU without a reboot... http://support.amd.com/us/kbarticles/Pages/73...pondingerrormessages.aspx Came across this while trying to find max temperature for the 6950. Still looking for that.
Crunching for humanity since 2007!
![]() |
||
|
|
mmstick
Senior Cruncher Joined: Aug 19, 2010 Post Count: 151 Status: Offline Project Badges:
|
There is no 'max temperature' besides a physical limit in which silicon melts, you can look that up, around 120C+, not like you could possibly get that hot. As for when does errors occur in software results, the higher your clock frequency, the lower the temperature threshold is for creating errors, but this is not an absolute constant, the temperature at which errors are induced are unique to your own graphics card. The reason why heat causes errors in high frequencies is because heat is a form of radiation noise, and noise makes it hard for 111000111 to be heard as 111000111 and not, say, 111010111.
|
||
|
|
Simplex0
Advanced Cruncher Sweden Joined: Aug 14, 2008 Post Count: 83 Status: Offline Project Badges:
|
Sometimes you need to run the Microsoft Installer Clean Up utility and select AMD APP SDK and uninstall it there manually as the catalyst setup will sometimes not uninstall/install it even if it is checked. If I have issues again, will give this a shot. Thanks. Is it only when running the HCC project you have this problem and not when running other GPU applications in BOINC ie. PHOEM@home or Milkyway@home? [Edit 1 times, last edit by Simplex0 at Oct 15, 2012 9:34:59 PM] |
||
|
|
Bearcat
Master Cruncher USA Joined: Jan 6, 2007 Post Count: 2803 Status: Offline Project Badges:
|
Only issues are with HHC GPU only. Am not crunching any other project at this time.
----------------------------------------
Crunching for humanity since 2007!
![]() |
||
|
|
Simplex0
Advanced Cruncher Sweden Joined: Aug 14, 2008 Post Count: 83 Status: Offline Project Badges:
|
Only issues are with HHC GPU only. Am not crunching any other project at this time. I'm just wondering if it can be a PSU related problem but if try to run milkyway or POEM and that works fine you can be sure that the problem it is not PSU related. |
||
|
|
Bearcat
Master Cruncher USA Joined: Jan 6, 2007 Post Count: 2803 Status: Offline Project Badges:
|
Only issues are with HHC GPU only. Am not crunching any other project at this time. I'm just wondering if it can be a PSU related problem but if try to run milkyway or POEM and that works fine you can be sure that the problem it is not PSU related. Currently using a Corsair 850 Gold rated PSU. Since it does work fine with milkyway, pretty sure its how WCG coded the application. Haven't run milkyway in a few months, just unticked HCC to switch over once my cache of HCC dries up just to check again. But since reinstalling CCC, everything seems to be back to normal except the lag. One interesting note is I read a comment somewhere here on these forums that since my card is double precision that its possible its not running at full speed. Have no clue at this time as to why and to many folks guessing quick fixes so will just let it go. Hopefully, WCG can refine the code in future projects. Am currently looking for a 7950 as it seems the 7000 series don't have many issues. Thanks for the response. Another note. Have read in another thread that folks are seeing better results shutting off hyperthreading. Just lowered from 22 threads and do see less lag. The lower I go, the less lag but still some at 18 threads. Might try removing hyperthreading just to see if it does work better. Note 2. Played around with the hyperthreading and turbo by shutting them off. A little better but not worth leaving these off. Back on. Note 3. Changed out to HDMI from display port. A noticable difference but still some lag. Since playing with the settings and trying different options, back to the driver kicking out. Have uninstalled and reinstalled different CCC packages but still screwed up. Have it all uninstalled now and running file check in hopes it finds any corruption. My 5670 on the other cruncher is doing fine. Wondering if double precision cards don't like HHC GPU seeing others with double precision are having issues vs 7700 and 7800 (single precision) don't seem to have issues from what I have read so far.
Crunching for humanity since 2007!
----------------------------------------![]() [Edit 2 times, last edit by Bearcat at Oct 16, 2012 7:25:28 PM] |
||
|
|
|