Are QMD's crashing my rig?

I changed my settings on the F@H program to allow big workunits and -advmethods -forceasm and my rig crashed overnight. The computer didn't shut off or reboot, the screen just went dead while everything else kept running. At first it looked like a monitor cable was disconected. I thought it was that until I tightened the cable, rebooted and checked the log and sure enough it had stopped working on the unit about 4 hours before I discovered the problem. I fired up F@H again it is started crunching away but an hour later when I checked it it had happened again. After reboot I read the log and it had been working for at least 45 minutes.

I have been running F@H for several days now on other WUs with no problems and I dont think it is stability issues because this rig is not overclocked. The only thing I can think of is the -forceasm. I'll try that soon,but I just wanted to check with you guys to see if there are any other known issues.

Rig specs:

p4 2.8c 478pin
1gb ram
intel d865gbf mobo

don't think gfx card matters but its a x800pro

It isn't -forceasm. All that does is force the core to use optimizations after a crash. No way it can cause a crash. I'd blow out the heat sink first and then run prime95 and memtest86. Which client are you running? If it's the Graphical client, there are known Open GL issues that cause crashes.

Yeah, I'll second what ChasR has posted. Especially about the graphical client. It definitely has known issues and from what little I know about the problem, there's no real easy way for Stanford to get around some of the problems either. If you are using the graphical client I would highly recommend that you d/l the console client and use it instead. If you really need to see the pretty pictures, then d/l EMIII and then d/l the chime plugin. The chime plugin will make real pretty representations of the wu you are working on.

Alright thanks for the advise. Strangly enough, I took the -forceasm off and it hasn't crashed again, but I'm not arguing cause and effect here. Also I have not brought up the display for more than a minute at a time, so that might be why it hasnt crashed. If I remember right, the display was up overnight when it did crash.

Can I stop work on the wu now and finish it with the console or do I need to finish it with the graphical client?

I am not familiar with moving WU's around with the graphical but i expect it should work ... Usually i just copy the folder called 'work', queue.dat and unitinfo.txt to the console directory. Keep a copy of those files to make sure it works ok. I guess it will want to download the core that was running it so I guess you should copy that over too ... FahCore_96.exe for QMD's.

When I do this between consoles it just picks up where it left off.

Good luck.