9401 fails on GM107 but not GK106 {Hopefully fixed}

[Breach]

When did you last try? You should not be getting these projects on FAH now.

[gaitskill]

Is there any way to quickly check that a card is failing

I have checked the Nvidia site and the Asus (my card) site and found none.

Now that video cards are being used as computers, there is a real need for something to load them and check for accuracy similar to the Intel Burn Test being used to check CPUs.

[Breach]

Plenty of tools to test for stability - FurMark, EVGA OC Scanner, Valley, Heaven benchmarks, 3D Mark...

[7im]

Folding@home's own FAHBench stress tester is a good one. And these on the fah web site: https://folding.stanford.edu/home/download-utilities/

[Breach]

Folding@home's own FAHBench stress tester is a good one. And these on the fah web site: https://folding.stanford.edu/home/download-utilities/

Speaking of which it seems http://www.fahbench.com/ is down...

[bruce]

Now that video cards are being used as computers, there is a real need for something to load them and check for accuracy similar to the Intel Burn Test being used to check CPUs.

Plenty of tools to test for stability - FurMark, EVGA OC Scanner, Valley, Heaven benchmarks, 3D Mark...

I, too, recommend FAHBench. There's nothing wrong with those other tests, but they include code that validates video components which FAH doesn't use, whereas FAHBench stresses the specific gpu sections that FAH does use.

[7im]

Folding@home's own FAHBench stress tester is a good one. And these on the fah web site: https://folding.stanford.edu/home/download-utilities/

Speaking of which it seems http://www.fahbench.com/ is down...

Use the source... https://github.com/SimTk/fahbench

[Breach]

I doubt most users here are exactly savvy compiling CPP code

[7im]

I doubt most users here are exactly savvy compiling CPP code

I contacted the dev for a copy of the Windows version so I could add it to (host from) the FAH web page if the fahbench site is not coming back. But he is very busy with the GPU and AS issues at the moment, so a response may be slow in coming.

[gaitskill]

Can you list the projects that were successful before separately from the projects which fail now? Are any projects on both lists?

Please confirm which drivers you're using and when they were last updated.
I think it fair to assume you did not change the gtx750ti and overclocking settings didn't change.

I have restarted FAH and it has completed one core 17 project 9201 and started on a second.
This with driver 344.11 and standard factory overclock.

I do not have a good record of which projects ran and which ones failed prior to this last string of all failures. What I have is segments of the log captured between reboots. The ten failures that I posted consisted of eight project 13000 and two project 13001, all core 17 and on driver 344.11. The prior group of ten failures was a similar mix of project 13000 and 13001. The next previous group of ten failures were all project 9406.

Before this, core 18 project 10475 successfully completed.The next project 10475 failed at 94%. and then again at 96%. Project 9406 failed on initiation six times. Project 10470 completed successfully.Then project 10473 failed.

This card had been installed in computer "A" and had a number of UNSTABLE_MACHINE errors.

In trying to isolate the problem, I switched video cards with computer "B",this machine, which was running FAH with out error. Machine "A" now runs without error and problems I now have are on machine "B".

Driver 334.94 was installed originally from the ASUS setup disk. Initiation of the failures was not caused by changing to other drivers. The changes were a futile attempt to stop the failures that had already started.

[james888]

For me, the only problem now is the lack of availability sometimes for core 17 wu's.