[ContextCapture] Windows timeout detection and recovery (TDR) error

What is GPU load while processig if using GPU-Z tool?

Theses are my results. Semms strange that so few memory of my GPU is used, or ?

What are maximums frequencies and GPU load you observe? Also what is most frequent limiting factor for GPU shown in PerfCap?


What I see from my observations is that GPU load is jumping from 0 to 100% while production is calculating, then when progress is at 18% GPU load drops to 0% but CPU load is 100%. Afterwards it again switches to GPU calculations and GPU load fluctuates from 0 to 100% till production finishes. Average GPU load for productions is 25%

Hi Pierre,

I had to stop the project it was stuck at 94% for 2 days... Yest I relaunched from 70% because before it was taking too long so I stopped.
Even relaunching I get the same problem. I also try upgrading to the newest version of CC but nothing changes unfortunately...

Best regards

Hi Oto,

yes I have similar observations using Z-GPU and the windows processus manager list of CPU use...

Do you have latest CC version v4.3.0.507?

On Windows server 2012 there is one more setting to increase DPC_WATCHDOG_VIOLATION timeout :

HKLM\SYSTEM\CurrentControlSet\Control\Session Manager\Kernel\
    DpcWatchdogPeriod=REG_DWORD 0003a980 (hexidecimal)
    DpcTimeout=REG_DWORD 00000000

On Windows server 2012 there is one more setting to increase DPC_WATCHDOG_VIOLATION timeout :

HKLM\SYSTEM\CurrentControlSet\Control\Session Manager\Kernel\
    DpcWatchdogPeriod=REG_DWORD 0003a980 (hexidecimal)
    DpcTimeout=REG_DWORD 00000000