I've been running 301.42 on my EVGA GTX 580 SC 1.5 for the past couple months with no problems and great performance. I run a dual monitor setup and commonly game with WoW running on one monitor (crafting) and SWTOR running on the other monitor (playing) with maxed out settings at around 60-100 FPS (in both games).  I've got a custom fan speed setup and usually stay below 70 degrees at full load...neither my video card nor my cpu is overclocked. Basically, I treat my system extremely carefully aside from pushing my performance limits within the stock settings. 
 
This morning, I woke the computer from sleep to find that my 3d performance had decreased dramatically. I'm now getting 30-50 FPS in SWTOR or WOW when playing each on individually and 20-30 FPS when playing the games simultaneously. Also interestingly, the GPU is not under full load and never gets above ~75% load. Previously, I was at full load just playing SWTOR alone. I tried updating to the newest drivers, re-seating my GPU, etc, with no luck. My other benchmarks have decreased dramatically as well; Metro 2033 is down to averaging about 30 FPS in DX9 and my Furmark score is down to 1155. l also noticed that during my Furmark run the clock was down-throttling even more than normal; normally it throttles to 399 but this morning it was throttling down all the way to 199 regularly. During the Furmark run, my temp never got above 67 degrees.
 
Does this sound like a power issue? The only thing thats odd is that I also tested Max Payne 3 and it hit 100% load easily and performed at its usual 150-200 fps with no obvious problems, so that makes it seem like a software-specific issue. This afternoon (when I get home from work, love going to work with a problem at home waiting to be fixed) I'm going to reset my config and roll back my system to a full backup I did this weekend to see if its software related. I'll also try rolling back drivers to 296.10. Has anyone seen anything like this occur before? I made no significant changes to my system last night except switching out my second monitor a couple of times to test another monitor. The card has performed rock-solid until this point and I've treated it with kids gloves since I got it two months ago.

ferenz


I've been running 301.42 on my EVGA GTX 580 SC 1.5 for the past couple months with no problems and great performance. I run a dual monitor setup and commonly game with WoW running on one monitor (crafting) and SWTOR running on the other monitor (playing) with maxed out settings at around 60-100 FPS (in both games).  I've got a custom fan speed setup and usually stay below 70 degrees at full load...neither my video card nor my cpu is overclocked. Basically, I treat my system extremely carefully aside from pushing my performance limits within the stock settings. 

This morning, I woke the computer from sleep to find that my 3d performance had decreased dramatically. I'm now getting 30-50 FPS in SWTOR or WOW when playing each on individually and 20-30 FPS when playing the games simultaneously. Also interestingly, the GPU is not under full load and never gets above ~75% load. Previously, I was at full load just playing SWTOR alone. I tried updating to the newest drivers, re-seating my GPU, etc, with no luck. My other benchmarks have decreased dramatically as well; Metro 2033 is down to averaging about 30 FPS in DX9 and my Furmark score is down to 1155. l also noticed that during my Furmark run the clock was down-throttling even more than normal; normally it throttles to 399 but this morning it was throttling down all the way to 199 regularly. During the Furmark run, my temp never got above 67 degrees.

Does this sound like a power issue? The only thing thats odd is that I also tested Max Payne 3 and it hit 100% load easily and performed at its usual 150-200 fps with no obvious problems, so that makes it seem like a software-specific issue. This afternoon (when I get home from work, love going to work with a problem at home waiting to be fixed) I'm going to reset my config and roll back my system to a full backup I did this weekend to see if its software related. I'll also try rolling back drivers to 296.10. Has anyone seen anything like this occur before? I made no significant changes to my system last night except switching out my second monitor a couple of times to test another monitor. The card has performed rock-solid until this point and I've treated it with kids gloves since I got it two months ago.

1st off, welcome to the forums...Have you tried the latest beta driver, 304.79??  The 301.xx driver has known issues on some factory OC'd GPU's causing downclocking.  The new beta driver fixed that issue.
Also, when you say "roll or rolling back" drivers, are you talking about the old way of going in the Device Mgr and rolling back drivers from there??  If so, that doesn't work well with the new driver sets.  If that's what you meant, I would recommend reading through this guide...http://www.evga.com/forums/tm.aspx?m=1174372.  Otherwise, we'll have to look for other explanations after you try the latest beta driver.

Thanks for the response. I installed the 304.79 driver this morning after the issues began and it did not resolve the problem; I installed it simply by choosing "Custom Installation" and doing a clean install, so if this was not the proper method of installation then maybe I can try completely cleaning out the drivers and reinstalling.
 
The odd thing is that I had no problem with the previous drivers before this morning; in fact, my system has been rock solid since I built it two months ago. As I said before, I normally run multiple game instances on dual monitors with almost no noticeable performance hit (100+ fps WoW on one monitor and 60+ fps SWTOR on the other with max settings in both games) and my card has never gotten higher than 87 degrees (before I set up my custom fan routine in afterburner). 
 
In regards to my "rolling back" statement, the first thing I'm planning to try this afternoon is a full restore of a full windows 7 system backup (300 gigs) that I did this weekend to see if its a software issue. I believe this should place me back into the 301.42 driver set. I was also going to try an actual roll-back (using the windows 7 uninstall/clean install method) to the 296.10 drivers to see if that would have any change).
 
The problem is, I'm not sure how it could be a software issue because I haven't made any significant software changes since earlier this week when it was running great. Last night I switched out a couple of monitors while the card was active (hot plugging); I know this isn't considered 100% safe but I've been doing it for a decade with no ill effects. Other than that, I didn't make any significant changes to my hardware or software. 
 
I'm running am ASRock P67 Extreme 4, i2500k, 16gb G-Skill RAM, and up until this point I've never come close to fully taxing this system since I put it together. All I can think is that its easy a software issue that I've overlooked or its a gpu/cpu/power issue that I'm going to have to do some testing/logging to diagnose. Hopefully its just a software problem I somehow caused without realizing and a full system restore will fix it.

One more question for any knowledgeable folks...what would be the best way to determine/prove that this is a hardware issue should I get to the point that I am considering sending the card in for warranty/service? Should I log detailed gpu-z info, run any specific benchmarks, etc?

ferenz


Thanks for the response. I installed the 304.79 driver this morning after the issues began and it did not resolve the problem; I installed it simply by choosing "Custom Installation" and doing a clean install, so if this was not the proper method of installation then maybe I can try completely cleaning out the drivers and reinstalling.

The odd thing is that I had no problem with the previous drivers before this morning; in fact, my system has been rock solid since I built it two months ago. As I said before, I normally run multiple game instances on dual monitors with almost no noticeable performance hit (100+ fps WoW on one monitor and 60+ fps SWTOR on the other with max settings in both games) and my card has never gotten higher than 87 degrees (before I set up my custom fan routine in afterburner). 

In regards to my "rolling back" statement, the first thing I'm planning to try this afternoon is a full restore of a full windows 7 system backup (300 gigs) that I did this weekend to see if its a software issue. I believe this should place me back into the 301.42 driver set. I was also going to try an actual roll-back (using the windows 7 uninstall/clean install method) to the 296.10 drivers to see if that would have any change).

The problem is, I'm not sure how it could be a software issue because I haven't made any significant software changes since earlier this week when it was running great. Last night I switched out a couple of monitors while the card was active (hot plugging); I know this isn't considered 100% safe but I've been doing it for a decade with no ill effects. Other than that, I didn't make any significant changes to my hardware or software. 

I'm running am ASRock P67 Extreme 4, i2500k, 16gb G-Skill RAM, and up until this point I've never come close to fully taxing this system since I put it together. All I can think is that its easy a software issue that I've overlooked or its a gpu/cpu/power issue that I'm going to have to do some testing/logging to diagnose. Hopefully its just a software problem I somehow caused without realizing and a full system restore will fix it.

I personally thought the 301 drivers were really solid at 1st also.  It seemed like it took awhile before they began acting up causing me issues.  I also read in here that others experienced the same thing.  I think it's worth a try to uninstall through Prog/features all Nvidia related items. Then boot into safe mode & do a complete wipe of all graphics & physx drivers using driver sweeper or similar.  Then reboot into windows and try the 304 beta driver again.  I think since you said you tried MP3 I think it was, and it played fine, it doesn't sound like a hardware issue yet in my opinion.  Just out of curiousity, what PSU are you using??  Keep us posted.
 

I'm using a cooler master gx 750 (can't post link due to forum restrictions)

Other than being incredibly, annoyingly loud, I haven't noticed any issues with power up until this point. I'll try a thorough driver cleaning before I do anything more drastic, thanks for the help.

ferenz


I'm using a cooler master gx 750 (can't post link due to forum restrictions)

Other than being incredibly, annoyingly loud, I haven't noticed any issues with power up until this point. I'll try a thorough driver cleaning before I do anything more drastic, thanks for the help.

NP.  Doesn't seem to be a lack of power then as long as it's not failing in some way.  But it's good to try and rule out software 1st before jumping to failing hardware, in my opinion anyway.  Especially since you said your system is all pretty new.  Don't suppose you have another GPU you could swap with if the clean driver install doesn't work?  Would be a pretty easy test to see if you get the same results with a different GPU.

So after some more investigation, I'm getting closer to figuring this out and would appreciate anyone's insight/opinion if you've got any ideas. 
 
Friday night, I restored a windows system image from before the issue began and it fixed the problem immediately. Afterwards, the only significant changes I made were to re-enable a custom fan profile in MSI Afterburner...I did not reboot the system after doing this. The system worked great all weekend, but on Monday I installed some Windows 7 updates that required a system reboot; afterwards, the issue returned. 
 
This morning, I formatted the hard drive and installed a fresh copy of Windows 7. For my first testing, my system consisted only of:
 
(1) Motherboard/Chipset Drivers
(2) NVidia 296.10 Drivers (Graphics Drivers/Physx Only)
(3) DirectX
 
The issue was not present and my performance was normal (~110 fps in SWTOR with full 99% GPU usage). I saved a backup image and then added all of the utilities I normally use:
 

 
I set up a custom fan profile again, tested everything, had no problems, and saved another backup. However, after doing some other work that required a reboot, I found that the issue returned. In retrospect, I also remembered that Wednesday morning of last week was the first time that I setup a custom fan profile, and I did not reboot the machine again until Thursday night (the issue first began on Friday morning). I'm restoring my last (pre-Precision) backup in order to test this, but my strong suspicion is that somehow setting up the custom fan profile is causing the video card performance problems. Has anyone ever heard of anything like this?

ferenz


So after some more investigation, I'm getting closer to figuring this out and would appreciate anyone's insight/opinion if you've got any ideas. 

Friday night, I restored a windows system image from before the issue began and it fixed the problem immediately. Afterwards, the only significant changes I made were to re-enable a custom fan profile in MSI Afterburner...I did not reboot the system after doing this. The system worked great all weekend, but on Monday I installed some Windows 7 updates that required a system reboot; afterwards, the issue returned. 

This morning, I formatted the hard drive and installed a fresh copy of Windows 7. For my first testing, my system consisted only of:

(1) Motherboard/Chipset Drivers
(2) NVidia 296.10 Drivers (Graphics Drivers/Physx Only)
(3) DirectX

The issue was not present and my performance was normal (~110 fps in SWTOR with full 99% GPU usage). I saved a backup image and then added all of the utilities I normally use:



I set up a custom fan profile again, tested everything, had no problems, and saved another backup. However, after doing some other work that required a reboot, I found that the issue returned. In retrospect, I also remembered that Wednesday morning of last week was the first time that I setup a custom fan profile, and I did not reboot the machine again until Thursday night (the issue first began on Friday morning). I'm restoring my last (pre-Precision) backup in order to test this, but my strong suspicion is that somehow setting up the custom fan profile is causing the video card performance problems. Has anyone ever heard of anything like this?

I've only heard of something like that when AB and Precision were running at the same time.

Well, after another 15-20 hours of troubleshooting I'm pretty much stumped. I thought that the issue was due to fan control profile; I've formatted and reinstalled Windows 7, and then incrementally installed the bare minimum of drivers, etc, in order to test to try to figure out the root issue. For the past two days I've been running a fresh OS without ever switching from the auto fan profile (which I hate because my temps sit at 87-88 with only 60% fan speed) for two days with no problems, so I thought that it was related to the fan control profile, but this afternoon I woke my computer from sleep to find that once again my performance in all 3d applications has been reduced by 50%...this time with literally NO software changes since this morning (and really since two days ago). Further, I'm getting minor shadow artifacts in SWTOR (cape shadow streaks black across screen). 
 
I'm pretty much resigned to RMA but now I have to deal with the choice between being without a video card (the one I just paid 500 bucks for three months ago) for a week or two or paying another 500 dollars to get a replacement using cross-ship and then waiting however long it takes to get a refund after my card is received/tested. And the thing that frustrates me the most is that apparently I'm the only one in the entire internet-capable world who has this issue, because I can't find a single other similar issue posted anywhere. It's uncanny, I can boot up a game to a character selection screen, watch it sit at 110 fps constantly for two mins, leave my computer, come back a couple of hours later, boot up to the same character selection screen, and now the fps has dropped to 60...I have no clue what could be causing this. 

ferenz


Well, after another 15-20 hours of troubleshooting I'm pretty much stumped. I thought that the issue was due to fan control profile; I've formatted and reinstalled Windows 7, and then incrementally installed the bare minimum of drivers, etc, in order to test to try to figure out the root issue. For the past two days I've been running a fresh OS without ever switching from the auto fan profile (which I hate because my temps sit at 87-88 with only 60% fan speed) for two days with no problems, so I thought that it was related to the fan control profile, but this afternoon I woke my computer from sleep to find that once again my performance in all 3d applications has been reduced by 50%...this time with literally NO software changes since this morning (and really since two days ago). Further, I'm getting minor shadow artifacts in SWTOR (cape shadow streaks black across screen). 

I'm pretty much resigned to RMA but now I have to deal with the choice between being without a video card (the one I just paid 500 bucks for three months ago) for a week or two or paying another 500 dollars to get a replacement using cross-ship and then waiting however long it takes to get a refund after my card is received/tested. And the thing that frustrates me the most is that apparently I'm the only one in the entire internet-capable world who has this issue, because I can't find a single other similar issue posted anywhere. It's uncanny, I can boot up a game to a character selection screen, watch it sit at 110 fps constantly for two mins, leave my computer, come back a couple of hours later, boot up to the same character selection screen, and now the fps has dropped to 60...I have no clue what could be causing this. 

I only have 1 other thought on this.  Try NOT letting your computer go to sleep.  If that means turning it off cause you won't be using it awhile, so be it.  Just to see if there's something going on after it comes out of sleep mode.  I can't really say much about the sleep mode thing because I've never used it on any of my builds.  I do with a laptop, but that's not like my gaming rigs.

Yeah, I'm going to see if I can get rid of the issue again (through system restore) and then test it with no sleep for awhile. The oddest thing is that it has characteristics that make it seem like it's definitely a software problem (doing a format/reinstall or a system restore sometimes fixes it) but also seems like it must be a hardware problem (happens randomly with no software change having taken place before the problem starts). I'm writing up a full summary of my tests/findings and posting it on my support ticket along with log data from precision, pics of artifacts in game, and benchmark results, so we'll see what happens. 

ferenz


Yeah, I'm going to see if I can get rid of the issue again (through system restore) and then test it with no sleep for awhile. The oddest thing is that it has characteristics that make it seem like it's definitely a software problem (doing a format/reinstall or a system restore sometimes fixes it) but also seems like it must be a hardware problem (happens randomly with no software change having taken place before the problem starts). I'm writing up a full summary of my tests/findings and posting it on my support ticket along with log data from precision, pics of artifacts in game, and benchmark results, so we'll see what happens. 

Keep us posted.  I'll be curious to hear what the results turn out to be.

Well, before I attempted to make any software changes yesterday I turned the machine off (after the performance drop had occurred) for awhile, removed the power cable, and discharged all remaining power by hitting the power button. After that, I turned the machine on to find that the issue was resolved (my performance was normal). I turned my fan profile back on, played fine all night, then turned the machine off. When I woke up this morning and booted the machine everything was fine. It might be completely related to some kind of issue with the machine under-performing after coming out of sleep (that can be a common problem, as I understand).
 
Before, when the problem occurred, I would reboot to attempt to fix it, but it would not have any effect. Now I'm wondering if perhaps rebooting does not clear whatever hardware/power/irq problem occurs after sleep and perhaps turning off the machine and doing a full discharge would have fixed it before (although I am almost sure I tried that once and it didn't work). 
 
Anyway, right now I'm running with a custom fan profile keeping my card under 70 degrees at full load (which I'm very happy about), so I'll keep testing it by avoiding sleep completely and let you know if the issue occurs again (or doesn't!). 

ferenz


Well, before I attempted to make any software changes yesterday I turned the machine off (after the performance drop had occurred) for awhile, removed the power cable, and discharged all remaining power by hitting the power button. After that, I turned the machine on to find that the issue was resolved (my performance was normal). I turned my fan profile back on, played fine all night, then turned the machine off. When I woke up this morning and booted the machine everything was fine. It might be completely related to some kind of issue with the machine under-performing after coming out of sleep (that can be a common problem, as I understand).

Before, when the problem occurred, I would reboot to attempt to fix it, but it would not have any effect. Now I'm wondering if perhaps rebooting does not clear whatever hardware/power/irq problem occurs after sleep and perhaps turning off the machine and doing a full discharge would have fixed it before (although I am almost sure I tried that once and it didn't work). 

Anyway, right now I'm running with a custom fan profile keeping my card under 70 degrees at full load (which I'm very happy about), so I'll keep testing it by avoiding sleep completely and let you know if the issue occurs again (or doesn't!). 

I hope it works for you.  I thought I remembered reading somewhere about issue's with sleep mode, but because I can't remember any details, I just thought i'd mention it to you as something to try out.

Resurrecting this thread because the issue returned and I wanted to see if anyone had any new ideas for what could be causing this.
 
I installed 306.23 WHQL drivers on Wednesday morning (up from 306.02) and played Borderlands 2 and Guild Wars 2 for awhile. In guild wars 2 I fluctuate between 80-140 FPS with my current graphics settings. Later that morning, I installed d3doverrider to force triple buffering and also set the evga precision frame limiter to 59 both globally and for individual games settings (was attempting to test a new setup for vsync with no input lag). Finally, I backed up my system to a disc image using Windows 7 Backup and Restore (did this remotely from work using LogMeIn). I left the machine running the rest of the day, but as far as I know the machine did not enter sleep mode since I've disabled this through Windows settings.
 
When I got home last night and logged into Guild Wars 2, I noticed immediately that my framerate had dropped to 25-50 in areas where I normally get 100-140 fps. Thinking that I had made a mistake with my frame limiter settings, I reset all frame limiter settings in precision to 0, disabled triple buffering forcing in d3doverride and removed it from win7 startup list, and powered down/booted up machine. The problem persisted and in all games tested (Borderlands 3, Guild Wars 2, SWTOR, WOW) my performance was reduced by over 50% across the board. I tried numerous reboots and cold boots, uninstalled 306.23 drivers and reinstalled 306.02, and nothing fixed the problem. Finally, I used windows system restore and went back to a restore point before I installed 306.23 and when the system rebooted the problem was gone and my performance was back to normal.
 
I'm still as confused by this as I was when it first started happening. If the issue is related to the driver level modifications from d3doverride and precision, then shouldn't it go back to normal when I remove those settings? Also, the issue doesn't seem to have any connection to any driver versions (it's happened across pretty much every set of drivers for the past 3 months). Finally, if it's hardware related and has to do with sleep settings, shouldn't it reset if I power down and cold boot the machine after draining power?
 
Nothing seems to make sense, and all I can do for now is make sure to have plenty of restore points to reset to once it starts happening and try to avoid doing any sort of driver level modifications or letting the computer go into any sort of sleep state. Why would resetting to a restore point work but resetting precision and cleaning drivers not work? This problem is so inconsistent that it's maddening to try to diagnose.

This is a crazy issue, which is related to your hardware not completely syncing up to communicate correctly. There is only two ways to solve the problem one of those options will cost you a lot of time and money, the other option will cost you time and frustration but no money.
 
#1 Keep rebooting your computer and testing your game till your performance is back to normal. e.g. reboot test fail, reboot test fail, reboot test pass.
 
#2 Start replacing your hardware with brand new hardware till you find the problem.

ferenz


Resurrecting this thread because the issue returned and I wanted to see if anyone had any new ideas for what could be causing this.

I installed 306.23 WHQL drivers on Wednesday morning (up from 306.02) and played Borderlands 2 and Guild Wars 2 for awhile. In guild wars 2 I fluctuate between 80-140 FPS with my current graphics settings. Later that morning, I installed d3doverrider to force triple buffering and also set the evga precision frame limiter to 59 both globally and for individual games settings (was attempting to test a new setup for vsync with no input lag). Finally, I backed up my system to a disc image using Windows 7 Backup and Restore (did this remotely from work using LogMeIn). I left the machine running the rest of the day, but as far as I know the machine did not enter sleep mode since I've disabled this through Windows settings.

When I got home last night and logged into Guild Wars 2, I noticed immediately that my framerate had dropped to 25-50 in areas where I normally get 100-140 fps. Thinking that I had made a mistake with my frame limiter settings, I reset all frame limiter settings in precision to 0, disabled triple buffering forcing in d3doverride and removed it from win7 startup list, and powered down/booted up machine. The problem persisted and in all games tested (Borderlands 3, Guild Wars 2, SWTOR, WOW) my performance was reduced by over 50% across the board. I tried numerous reboots and cold boots, uninstalled 306.23 drivers and reinstalled 306.02, and nothing fixed the problem. Finally, I used windows system restore and went back to a restore point before I installed 306.23 and when the system rebooted the problem was gone and my performance was back to normal.

I'm still as confused by this as I was when it first started happening. If the issue is related to the driver level modifications from d3doverride and precision, then shouldn't it go back to normal when I remove those settings? Also, the issue doesn't seem to have any connection to any driver versions (it's happened across pretty much every set of drivers for the past 3 months). Finally, if it's hardware related and has to do with sleep settings, shouldn't it reset if I power down and cold boot the machine after draining power?

Nothing seems to make sense, and all I can do for now is make sure to have plenty of restore points to reset to once it starts happening and try to avoid doing any sort of driver level modifications or letting the computer go into any sort of sleep state. Why would resetting to a restore point work but resetting precision and cleaning drivers not work? This problem is so inconsistent that it's maddening to try to diagnose.

After reading this (your) post, I get the impression you had been running good until you installed d3doverrider??  Just out of curiousity, why are you using d3doverrider to force triple buffering when you can do this right in the NVCPL for any specific game you want it for??  Maybe you're having some kind of conflict between the graphics driver, d3doverrider, and Precision.  So then when you went back to a restore point, d3doverrider no longer was an issue, and you were back running fine.
 

ATTN I was using a Cooler Master GX-750W and I started having problems with performance decreases... I decided to probe the +12v with a DMM, and discovered it at 12.56v! It should should not exceed 12.4v. Over time, a +12v that is outputting too high or too low can damage your graphics card and other hardware. My GX-750W damaged one of my GTX 460's, actually.
 
You may want to check it yourself if you have a DMM, there are guides you can find on Google that tell you how to check the voltages. Red = 5v, orange = 3.3v, and yellow = 12v. black = ground.
 
I do not trust Cooler Master's PSU's. Their cases and HSF's are great, but I would never buy a PSU from them again.
I'm not saying this is what the problem is, it may or may not be, but you may want to get it checked asap. If you don't want to do it yourself, any computer repair place can look at it for you.
 
If your +12v is out of spec, I would recommend not using that power supply anymore, and you would have to RMA it to Cooler Master.