OP
- Joined
- Apr 29, 2002
- Thread Starter
- #21
So far, so good. My PC is also doubles as a 700W space heater.
-
Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!
SOLVED GPU folding issue = hanging & lost WUs
- Thread starter Audioaficionado
- Start date
- Joined
- Jul 1, 2004
- Location
- 1 Infinite Loop
That is sooooo true. Spring is coming fast. Warmer temps on the horizon. I'm going to have to accelerate the upgrade to GTX 970 cards to cut down on the watts and heat.
OP
- Joined
- Apr 29, 2002
- Thread Starter
- #23
Well that was annoying. Just had a BSOD while posting everything was going great
Max vcore = 1238mv. Bumped it up again to 1244mv and it started artifacting.
So what's the going PPD for a GTX 970 and the W/PPD?
Max vcore = 1238mv. Bumped it up again to 1244mv and it started artifacting.
So what's the going PPD for a GTX 970 and the W/PPD?
OP
- Joined
- Apr 29, 2002
- Thread Starter
- #24
What the heck is up with GPU 1? That card also just lost a WU after a BSOD.
- Joined
- Sep 7, 2013
Looks as if F@H is trying to use the GPU, but it won't play nice for some reason...
Very odd that the GPU artifacts when you feed it more volts, never heard of that before...
Would you be able to put that GPU in another rig and try folding on it? If it's fine, then we know the card is still okay.
Very odd that the GPU artifacts when you feed it more volts, never heard of that before...
Would you be able to put that GPU in another rig and try folding on it? If it's fine, then we know the card is still okay.
- Joined
- Dec 25, 2004
Drop the core speed a bit more. Shut down the FAH client, kill off any FAH processes (core17, etc) after you've shut it off. Restart the client. When you see that in afterburner, the client has failed on just that core.
Sorry they are being frustrating for you.
Sorry they are being frustrating for you.
- Joined
- Jul 1, 2004
- Location
- 1 Infinite Loop
260-270K (p9201 and p9411) per day for 350W full bore. Drops considerably on p8018, etc. Works for me.
OP
- Joined
- Apr 29, 2002
- Thread Starter
- #28
260-270K (p9201 and p9411) per day for 350W full bore. Drops considerably on p8018, etc. Works for me.
Is this just the GTX 970 card or the system + card?
Both my cards are doing ~150kPPD for ~300W just for the cards. I imagine the total system cost is ~500W.
Looks as if F@H is trying to use the GPU, but it won't play nice for some reason...
Very odd that the GPU artifacts when you feed it more volts, never heard of that before...
Would you be able to put that GPU in another rig and try folding on it? If it's fine, then we know the card is still okay.
The WU is still progressing, but not as efficiently. I'll see if it finishes or gets corrupted. I think these cards are already OCed from the factory, so there's not much headroom to play with.
Not frustrated at all. This is expected when you are pushing hardware to the limits and part of the overall fun. We are Overclockers.com are we not?
I'll try lowering the core speed a tad bit. I'm real close to 24/7 stability and only lost a few WUs so far. Can't complain about your selling price
OP
- Joined
- Apr 29, 2002
- Thread Starter
- #29
FAH off & GPU idling after PC reboot
GPU 1 is also connected to 1080P monitor (although I don't remember it ever doing this)
GPU 1 is also connected to 1080P monitor (although I don't remember it ever doing this)
- Joined
- Jul 1, 2004
- Location
- 1 Infinite Loop
System plus a single GTX 970 FTW OCed card. Looking at my other system with two GTX 970 Superclocked cards (only factory overclocked) it is pulling 389W (working p8018 and p9201) according to the trusty old Kill-A-Watt P3.
OP
- Joined
- Apr 29, 2002
- Thread Starter
- #31
Looks like I'll have to upgrade pretty soon. I've gotten addicted to the higher PPD, but I have to cut power usage. Figure a couple GTX 970 cards should be good for ~500+kPPD for less power usage.
OP
- Joined
- Apr 29, 2002
- Thread Starter
- #32
GPU_FAH working and no clients/core/wu stalling.
The mouse and other input devices still work on the screen when the WU is stalled. So maybe it's something to do with the client/core that's stalling out.
The mouse and other input devices still work on the screen when the WU is stalled. So maybe it's something to do with the client/core that's stalling out.
- Joined
- Jul 1, 2004
- Location
- 1 Infinite Loop
Yes. Unless it happens to be two p8018 units running for 80K per day and bumps it up to 404W. Peaks and valleys.
OP
- Joined
- Apr 29, 2002
- Thread Starter
- #34
Yes. Unless it happens to be two p8018 units running for 80K per day and bumps it up to 404W. Peaks and valleys.
Sounds like a real sucky WU. Too bad I can't run AMD cards in Linux to avoid them.
- Joined
- Jul 1, 2004
- Location
- 1 Infinite Loop
OP
- Joined
- Apr 29, 2002
- Thread Starter
- #36
Well I was at my screen and noticed brief pixelation and the mouse pointer froze. Then after a few seconds the monitor went black a few seconds and then came back on. Got a popup notice that the graphics driver had crashed, but had recovered. No BSOD . Looking at the Afterburner graphes, GPU1 was doing its usual jaggies which means it's idling. So instead of quitting both FAH clients, I just paused FAH. Gave it a minute and restarted. Both clients started back up. The stalled client started from the last checkpoint file and finished a few hours later.
So it looks like something is causing a driver crash, but the good news is the WUs don't have to be nuked to start over. Check points can save them after a soft restart.
. Looking at the Afterburner graphes, GPU1 was doing its usual jaggies which means it's idling. So instead of quitting both FAH clients, I just paused FAH. Gave it a minute and restarted. Both clients started back up. The stalled client started from the last checkpoint file and finished a few hours later.So it looks like something is causing a driver crash, but the good news is the WUs don't have to be nuked to start over. Check points can save them after a soft restart.
OP
- Joined
- Apr 29, 2002
- Thread Starter
- #37
Things have been running very smoothly (knock on wood). EOC stats graph is still climbing and I've progressed to the T32 top 20 producers for now. Maybe the top 20 ain't what it used to be with all the heavy hitters we've lost post -bigadv, but I'm happy to be there just the same. Summer heat is coming. I'll see how these cards behave when they get up into the 80°C range. Until then, WTF T32
- Joined
- Sep 7, 2013
That's the best news we could hear!
Keep on folding!!!
