hashcat Forum

One gtx 1080 running alone under windows 7/64 latest nvidia drivers, and tested with hashcat 3.1 and also
with 3.4, 1200 PSU, do a normal work if a simple dictionary attack is done.
Once it's given a rule or brute force attack, not much after, stop with the following error:

ERROR: clEnqueueWriteBuffer(): CL_OUT_OF_RESOURCES

Msiafterburner is running, showing the temperature, but after this, and the nvidia driver recover, the
temperatue is 0 (zero).

If try to repeat the command line, which is very simple, the following message is displayed:

ERROR: No devices found / left

It is possible to run FurMark immediately, or other benchmark/stress gpu software like Unigine Valley Benchmark
with no signal of problems.

Only can have the card "available" again to run hashcat after restarting the computer.
I suspect this might be an hardware problem with the graphics card but until now I can't find a way to confirm
this by running anything else than hashcat ! Already used the same card in two different computers with the same problem.

Have more identical cards that run with no problems at all which makes me believe that the card, somehow, is faulty.
Any sugestions about how to confirm that the card isn't stable other than using hashcat ?
Thank you !

(04-12-2017, 11:43 PM)epixoip Wrote: [ -> ]https://hashcat.net/wiki/doku.php?id=timeout_patch

Thank you for your support epixoip.
Just found that topic, followed the instructions, and now I'm doing the same run but, initialy, with -w 1
and see how it behaves. If it ends with no problems will do it again changing to -w 2 and then -w 3.
I still don't get why the other cards, 100% identical, do hard work with no problems at all.

OK. It looks like this is solved !
After applying the patch, did a brute force run, 21 minutes with no problem at all.
This use to last just a few minutes, not so long !
If all good from now on, no more reply from me to this topic.
Thanks to the forum search archive, and epixoip.