BSOD only when playing PCSX2?
#1
I keep on getting BSOD when I play on PCSX2. It only happens after playing for a bit and not when the game starts up or anything. I haven't really noticed this before, and I am playing Dot Hack Infection right now. I have been playing with the frame limiter off, which is not something that I always do, but, for the sake of time, have been although I don't remember having an issue with having the frame limiter off in the past. I don't know if it's the game or what. I get BSOD with both 0.9.7 and 0.9.8. The most recent game I played before this was Xenosaga Episode III, and it worked flawlessly. Is it a problem with the game? Perhaps the fact that I'm using no frame limiter? Or maybe my processor is getting burnt out from overclocking? It's been deemed a stable overclock by my standards. I'm using a an AMD 965 Phenom II x4 processor, overclocked to 3.9ghz with a cooler master hyper 212 plus. Any help would be appreciated.
Reply

Sponsored links

#2
i would start by monitoring your temperatures etc.... PCSX2 is a very good stress test and is showing your overclock is unstable for some reason or other.
[Image: ref-sig-anim.gif]

Reply
#3
With the framelimiter off, you are definitely taxing your hardware more. But this shouldn't be giving you BSODs, unless the appropriate stress testing you did (or didn't) do resulted in BSODs, as well.

If it's not the temps being too high (you should be aware of them), it may need more voltage. No offense, but it may be time to question your standards for deeming an OC stable. Try Prime95 for hours, or several x20 passes of Intel Burn Test. Wink
Reply
#4
(05-07-2011, 02:08 AM)Rezard Wrote: With the framelimiter off, you are definitely taxing your hardware more. But this shouldn't be giving you BSODs, unless the appropriate stress testing you did (or didn't) do resulted in BSODs, as well.

If it's not the temps being too high (you should be aware of them), it may need more voltage. No offense, but it may be time to question your standards for deeming an OC stable. Try Prime95 for hours, or several x20 passes of Intel Burn Test. Wink

Haha it's definitely possible that my overclock isn't 100% stable. It's just never had an issue before, and it's been overclocked for over a year now. Who knows? I've only known how to overclock since last year, and I'm still not that good at it.

Using CPU-Z, CPUID Hardware Monitor and Linx, I tested out my overclock way back when last year, messed around with the multiplier and voltage until it passed 20 iterations just fine at 3.9 ghz. Nice. Simple. It was all fine, and, if I remember correctly, my comp idled at around 32C, and, on full stress, it went up to about 60C. But my computer with the frame limiter off was having the CPU reach 59C, but even with PCSX2 having the frame limiter on, the CPU temp wasn't dropping any lower than 54C. With the program completely shut down and just having my browser up, the CPU goes down to 50C. Those numbers do seem a little high. Only if I could afford to upgrade! Tongue

Take that information for what it's worth. What sounds good?
Reply
#5
Have you ever replaced the thermal paste? That 50c is pretty hot, if it's not doing much.

On the 20 passes, I've passed several before, only to eventually have an error. It doesn't always show right away. It can even get by for a long time being just shy of stability, but it also can eventually show.

An example would be if a clock needs 1.36v to be %100 stable, but you have it at 1.35v. You may get away with it for, actually, a really long time. Eventually, it will show. Especially if you introduce it to PCSX2, an app that can stress hardware to the likes of a stability test.
Reply
#6
(05-07-2011, 04:08 AM)Rezard Wrote: Have you ever replaced the thermal paste? That 50c is pretty hot, if it's not doing much.

On the 20 passes, I've passed several before, only to eventually have an error. It doesn't always show right away. It can even get by for a long time being just shy of stability, but it also can eventually show.

An example would be if a clock needs 1.36v to be %100 stable, but you have it at 1.35v. You may get away with it for, actually, a really long time. Eventually, it will show. Especially if you introduce it to PCSX2, an app that can stress hardware to the likes of a stability test.

I was thinking about doing that. Hopefully I can find mine or borrow someone's thermal paste. I installed the Hyper 212 myself, so it was replaced at that time, and I was pretty careful with it too. Which sounds more likely to be the problem? Heat or too low of voltage? Or is it something different altogether? It definitely only happens when the computer is under stress, so I'm assuming that it has something to do with the overclock.
Reply
#7
It most likely is the OC. You could try running at stock and see if the problem persists. If it doesn't, then your settings for that OC are probably off in some way. If the temp isn't going any higher than ~60c when the BSOD occurs, then it's not likely that heat is causing this. I can't guarantee it's then the Vcore, since there's more to OCing than that, ya know? It's a common culprit, that's for sure.

If your CPU is running 50c in that circumstance, then I'm guessing your paste may have seen better days. The stuff that comes with a CM Hyper 212+ is pretty decent, but no paste lasts forever. Again, if your temps stay 60c or under, that's not causing the BSOD.
Reply
#8
(05-07-2011, 04:55 AM)Rezard Wrote: It most likely is the OC. You could try running at stock and see if the problem persists. If it doesn't, then your settings for that OC are probably off in some way. If the temp isn't going any higher than ~60c when the BSOD occurs, then it's not likely that heat is causing this. I can't guarantee it's then the Vcore, since there's more to OCing than that, ya know? It's a common culprit, that's for sure.

If your CPU is running 50c in that circumstance, then I'm guessing your paste may have seen better days. The stuff that comes with a CM Hyper 212+ is pretty decent, but no paste lasts forever. Again, if your temps stay 60c or under, that's not causing the BSOD.

I used that arctic silver stuff instead of the stuff that came with the hyper 212. Hopefully I can find either that or the original thermal paste that came with the hyper 212 somewhere in my house! Tongue But anyhow, I think I'm going to reapply the thermal paste and try bumping up the voltage a smidge and see if either one helps. Btw, how often should thermal paste be replaced? It's been over a year.
Reply
#9
Yeah, definitely at least once within a year. I would think you'd need to replace it more often if you're using it on a overclocked CPU, too. A sure sign that it needs replacing would be hieghtened temps, but I would try to do it sooner (like, say every 6 months).

BTW - I think the stuff that came with your CM Hyper 212+ is better than Arctic Silver 5.
Reply
#10
Ok, so the replacement of the thermal compound was a huge success . . . at least in dropping my processor temperature down a good 10+C. I used the Cooler Master thermal paste like you suggested. Let's hope that the game will run without any crashes anymore. If it does, I'm going to bump up the voltage a smidgen.

Edit: Ok, so the processor temp still goes up to about 55C MAX when the frame limiter is off although the temp stays more around 52ish when the frame limiter is off, and I've been playing for over an hour, yet no crashes have happened yet. I have NOT changed the voltage.
Reply




Users browsing this thread: 1 Guest(s)