PC stuttering and freezing, requiring manual restart

[jktstance]

Hi all. I have an ASUS P6T and it uses the RealTek onboard audio hardware. However, I've been having some problems with it. In some games, including some heavy ones like Crysis and others like Dungeon Fighter Online, there will be periods of entire stuttering, both of sound and graphics. Sometimes I might get flickering in some graphics too. Steadily it will get worse and worse until eventually the entire PC freezes. The screen is frozen, the audio loops the same millisecond of sound, and nothing works. I have to cut the power.

I have a sneaking suspicion it has to do with the RealTek system, even though this problem affects my whole system. It doesn't happen all that often, but it's enough for me to consider buying a replacement part. The problem also might be indicative of not enough power, but that shouldn't be an issue as I have a very large PSU.

I recall having a similar problem with my previous system, which also used RealTek onboard audio (MSI K8T Neo if I recall correctly).

I have been updating the sound drivers but to no avail. Does anyone know what might be the problem?

 

[T_T]

Generally, problems like this (freezing/audio loops) are problems with RAM. I had a similar problem, it turned out to be a defective GPU. If you want to be sure it's not your onboard audio, disable it, then run the game again.

You should also use MemTest86+ to test your RAM.

Things to look for:

1. GPU temps - Download and install SpeedFan, GPU-Z, or HW Monitor to find out what the temps are. Ideally, you want this temp to stay below 60 C on load; however, depending on the ambient temp, 65 C isn't too bad. Any time you keep your GPU temps at 70 C or higher, you risk damaging the GPU.

2. CPU temps - Similar to GPU temps, use HW Monitor or SpeedFan to determine the CPU temps. This too, should stay below 60 C.

3. RAM errors - MemTest86+ will determine if there is a problem. Keep in mind that this test is independent of any OS, so don't worry about compatability. Also, test one stick of RAM at a time. Run the test for about 8-10 "passes".

 

[jktstance]

I'll try those. This is hard to test because the stuttering and freezing occurs pretty rarely.

 

[T_T]

I know how you feel. My experience like this was also very random. I tested RAM using MT86+4.0 (at the time) for 8+ hours; RAM was ok. Stressed my CPU using Prime 95. This ran for more than an hour, still no freezing. I even tested my GPU using FurMark, which does similar testing to the GPU that Prime95 does to the CPU. To my dismay, all tests were fine. However, what made me really believe that it was a GPU problem was the fact that upgrading or downgrading the driver would give me the same or worse results.

Now, I'm not saying that your problem is conclusively GPU, just throwing out suggestions that seem more relative as the cause of the freezing.

If there is a Window(ed) Mode to your games, I suggest you try using that while running the temp monitoring software in the background. FurMark, if you wanted to use it, will use your entire display, I.E. no adjustable window, but it will also give your GPU temp. I've seen my (current) GPU reach as high as 85 C while under load of FurMark, but I live in a warm climate.

 

[jktstance]

It seems that stress is no factor in my case. I can run Crysis at full graphics and it performs well. Sometimes it'll stutter, but not often. Most of the time, DFO (a 640x480 game) runs fine, but occasionally it has that catastrophic stuttering issue.

I'm in the process of making a memtest86+ bootable ZIP drive (too lazy to hook up my CDRW drive).

 

[T_T]

Indeed, but my suggestion wasn't to see if your system could handle the load, but rather to see what the temps were at the time of freezing. You could have intermittent failure in one or more of your fans (Case, Heatsink, GPU, PSU).

Post back with the results of MT86+ and any other software test results. I'm very curious as to the temps your system runs at when it is normal, and when it freezes.

By the way, this freezing, does it occur only when playing a game?

 

[jktstance]

Indeed, but my suggestion wasn't to see if your system could handle the load, but rather to see what the temps were at the time of freezing. You could have intermittent failure in one or more of your fans (Case, Heatsink, GPU, PSU).

Post back with the results of MT86+ and any other software test results. I'm very curious as to the temps your system runs at when it is normal, and when it freezes.

By the way, this freezing, does it occur only when playing a game?

Yes, T_T, it only seems to happen when playing a game. I'm thinking maybe the change in resolution when I open a fullscreen game might have something to do with it.

MT86+ finished with no errors. I've been monitoring my system with GPU-Z now. No freezing last night, but my GPU was running at a steady 70-73 C in game and somewhere in the 60s out of game. GPU fan was kept at around 50-60% top speed.

My GPU is a GeForce 285, the OC edition from BFG.

Now it's just the waiting game. I'll post back with figures when my system freezes, but it could take up to a week for that to happen. Thanks for the help thus far.

 

[jktstance]

A quick update.

I installed Riva tuner and I increased my GPU fan speed to 65%. This keeps the GPU at around 60-62. I have noticed less stuttering (though occasionally there is some) and no crashes. I'm a bit confused about this because when I was using the stock card settings, it would automatically increase the fan speed as needed. It was at around 73 degrees before, which isn't really that hot, certainly nothing that would cause freezing, I would imagine.

I have been running into a separate problem. The mouse will freeze for a second and I hear the Windows chime that a USB device was disconnected, then it would resume. I have a USB mouse on Windows 7.

Apparently this is a known issue with no fix from Microsoft: http://social.answers.microsoft.com/Forums/en-US/w7performance/thread/01655e0a-1489-44e0-ad75-e9ee5486d91e

 

[T_T]

In regards to the fan throttling, the reason why the GPU can get hotter is because the fan speed increase isn't triggered until the GPU reaches a factory specified temp. With your fans set manually at 65%, you're stopping the card from reaching the higher temps because of constant cooling.

I'll reply back about the mouse issue after i've read through the forum in the link you provided. But off the top of my head, the only real suggestions I have are reinstalling ALL of the software for the mouse or using USB to PS/2 adaptor/converter.

 

[jktstance]

In regards to the fan throttling, the reason why the GPU can get hotter is because the fan speed increase isn't triggered until the GPU reaches a factory specified temp. With your fans set manually at 65%, you're stopping the card from reaching the higher temps because of constant cooling.

I'll reply back about the mouse issue after i've read through the forum in the link you provided. But off the top of my head, the only real suggestions I have are reinstalling ALL of the software for the mouse or using USB to PS/2 adaptor/converter.

Hi T_T,

I understand about the temperature issue. I'm just surprised that 73 degrees is too hot. I'm sure it goes much higher than that in Crysis (at which point the GPU factory limit is reached and the fan speed is increased). If indeed GPU temperature is the culprit behind my stuttering and freezing, then it seems 73 wouldn't be so hot to cause problems.

That being said, I am still experiencing some stuttering here and there, but I haven't had a full system lock yet.

The mouse issue is a new one. It's only been happening for the past week or so. However, it's gotten REALLY bad lately. I'm going to be updating all of my motherboard drivers and BIOS. The mouse I have (Logitech MX510) does not have any driver software for Windows 7).

I ran the latency check program from http://www.thesycon.de/deu/latency_check.shtml. When the mouse stutters, the latency jumps to around 500 microseconds, then it goes back to it's normal <100 microseconds. I didn't see any jumps in the thousands like some of the screenshots were reporting.

The stuttering seems to happen the most during a game. I suspect it has to do with high internet communications. When I close the game, I get the occasional stutter, but nothing as bad as before.

Last resort with that is to try a USB -> PS/2 converter. I can't seem to find those online, though there are plenty of PS/2 -> USB converters.

Of course, the VERY last resort is to go back to Vista

 

[T_T]

The 285 OC is nice card and should handle Crysis, but from what I've read, Crysis is a very GPU-intensive game. That being said, try the following:

1. Resolution in game matches (or lower than) your monitor's native resolution
2. Turn down AA, Vsync, and triple buffering)
3. Turn down Shadow effects

See if any of this reduces or removes the graphical stuttering you're experiencing.


As for the adaptor, it should look like this:

and you can get it from http://www.dixmaxx.com search for "ps/2 converter".

 

[jktstance]

The 285 OC is nice card and should handle Crysis, but from what I've read, Crysis is a very GPU-intensive game. That being said, try the following:

1. Resolution in game matches (or lower than) your monitor's native resolution
2. Turn down AA, Vsync, and triple buffering)
3. Turn down Shadow effects

See if any of this reduces or removes the graphical stuttering you're experiencing.


As for the adaptor, it should look like this:
http://www.dixmaxx.com/images/dix_pccable0011.jpg


and you can get it from http://www.dixmaxx.com search for "ps/2 converter".

I see I just suck at finding stuff online...

I think I wasn't clear before. I meant to say that I have run Crysis in the past, max everything, with no problems at all. During that time, I wouldn't be surprised if the GPU exceeded 73 degrees and the fan speed increased.

The game I am currently playing is nowhere near as GPU-intensive (Vindictus) and the GPU sits at 73 degrees if I don't manually increase the fan speed), and yet I am getting stuttering and freezing. I have also experienced this in League of Legends, which is even less GPU-heavy, but I don't have any temperature readings for that at the moment.

That's why I think this issue is not related entirely to GPU heat, though manually upping the GPU fan speed seems to be showing something else. To be honest, Vindictus has HORRIBLE netcode and maybe that's causing some sort of problem with my network adaptor. LoL is pretty good, though.

Hopefully when I go through all of the motherboard updates things will smooth out. First time BIOS update here!

 

[jktstance]

I have some interesting news on this.

I ran event viewer and I was getting periodic messages about bad blocks in the hard drive. I scheduled and ran chkdsk -f -r. When that completed (I didn't see the results as I was away from the computer all day then), my computer doesn't seem to be doing the weird USB mouse disconnect bug. I do have a USB -> PS/2 adaptor now just in case.

However, my computer did have an episode of stuttering again. This time I was able to start up DPC latency checker to see what was going on. During the episodes of stuttering, my latency was quite high, well over 10,000 us. So while this was happening I went to device manager and I disabled various devices that didn't require a restart (things like the floppy drive). I got through a few devices until I accidentally pressed OK to restart the machine. DOH!

However, I did not reach the USB controllers in my device-disabling frenzy earlier, so it's possible they might be clogging my PCs internal communications.

The stuttering has not repeated itself for a few days now.

Also, I should point out that in my game, Vindictus, there is sometimes some graphical tearing or things like the sky will flicker. Also, rarely some objects like trees will be replaced with their 3D model, as if the textures were not loading. I always attributed this behavior to poor coding, as Vindictus seems to be rather buggy (but oh-so-addictive), but I suppose it might be indicative of something gone awry in my system.

 

[T_T]

I'm really leaning toward a failing GPU now. The only reason why i'm not positive is because a failing GPU would show itself in the OS, too. If the graphical failures only occur in Vindictus, then it is likely the problem of the game, not your GPU. On the other hand, what was the result of stability testing your GPU using FurMark?

 

[jktstance]

Hi. I never heard of FurMark until now, but I downloaded it (and the latest GPU drivers) and I let it run for a bit. I ran it using my stock fan controls. I was getting around 87 fps on a 1280x1024 screen of that rotation fur ring whatever. The GPU temperature went up to 88-89 degrees, which is where it seemed to hover at. I let it run for a few minutes there, then shut it off. Everything seemed very stable.

The GPU fan speed didn't start to increase until the temperature reached around 70 degrees, at which point it increased all the way to 100%.

I have a hunch that the system freezes and mouse stutters are somehow related, though it's difficult to test as I don't have a reliable way to create these problems. And they don't seem to appear simultaneously.

I'll try running Vindictus tonight without manually increasing my GPU fan speed. The GPU tends go to 70-73 degrees and the GPU sits at around 40% speed (it's idle rate). As for the mouse freezing, that hasn't happened for 5 days or so...

 

[bagelemage]

hi there. did you solve the stuttering then freeze issue?

i am having the exact same problem, only with games!

would love any hints

thanks

 

[jktstance]

Wow, late reply!

The issue for me is DPC latency. It's a really nasty problem that's almost impossible to track down. It's most likely caused by hardware or drivers and I've been unable to figure out what's going on. Even a ticket with Microsoft hasn't helped.

I used this program to figure out it was DPC issues. Just start it when the PC is stuttering: http://www.thesycon.de/deu/latency_check.shtml

I've tried using event monitors and such, but no such luck. I have a sneaking suspicion it's the sound card, because RealTek drivers have always given me problems, but I can't disable the device during a stuttering sessions without requiring a system reboot.

Basically, the only thing I can do is if my system is stuttering due to DPC latency, I have to restart and hope whatever set it off doesn't happen again. It tends to only start happening in games. Closing the game will result in the DPC latency gradually getting better, but it'll still be problematic.

EXTREMELY ANNOYING. No one seems to be able to help.

 

[bagelemage]

alternatively, could the DPC latency occuring during the crash be a result of the crashing system vs the cause?

i find it strange that it happens only during games, and even more so that is no longer happens if I reboot the system.
why would the reboot prevent the stutter/freeze from happening? vice versa: why does the problem happen only after a power-on from cold?

have you changed your graphic card during this time period?

i even formatted my system and reinstalled... still happening

 

[jktstance]

In my case, no. If I don't remove the system load (the game, in this case), the stuttering will eventually get worse to the point of a slideshow. The entire system stutters, not just video/audio. Mouse inputs, for example, are sporadic. Eventually that will lead to a BSOD.

Because the problem seems to reset itself on a system restart, as well as several full installs of Windows, from Vista up to 7, I'm pretty sure it's a hardware or driver issue for me. I've installed newer versions of most of my drivers as well as a BIOS update, to no avail.

As T_T said earlier in the thread, it could very well be a bad video card. I haven't changed mine, but it's starting to show it's age with Guild Wars 2. I've also recently overclocked my CPU, but system stability hasn't changed at all from that.

 

[bagelemage]

cool

i might change my card in the near future to a GTX 580 from 480

we'll see what happens