System crashing (BSOD) on Crysis 3 (RESOLVED?)

[marmasatt]

Update 5/19/14

This is insane. All of these BSOD's were corruption related it appears? I did a Windows 8 "upgrade" and that seemed to preserved the corruption. However, when I did a fresh reinstallation of the OS, everything is fine. Unbelievable, MS............

Where do I start troubleshooting this issue? I am running a 2GB GTX 460, 2500K, 8 gigs memory, 600 Watt PSU.

I used the Geforece experience to set my settings for optimal performance, but I figured my machine couldn't keep up so I began lowering stuff. Now I've lowered stuff all over the place in this game (Crysis 3) and I can't stop from crashing after about 10/15 minutes of playing. It's not doing it in other games but I'm not playing alot of taxing stuff right now either.

Is this definitely an overheating thing? Ran a few stress tests and it's not bsod'ing. Temps during stress were 65 degrees w/ fan running at 46%. (2400RPM). Should I turn the fans up? They already get pretty loud during the game itself. I think that would annoy me - but not more than the game crashing every 10 minutes. :\ Running the latest 337.5 Nvidia drivers from 4/7.

Thanks.

 

[lehtv]

What is the BSOD error code?

 

[marmasatt]

This is one of them: dxgmms1.sys bsod page_fault_in_nonpaged_area

Most recent anyway.

A little hotter too. Was about 70 degrees (still doesn't seem that bad) with fans spinning up to 66% - And RPMs like 3500.

 

[Makaveli]

This is one of them: dxgmms1.sys bsod page_fault_in_nonpaged_area

Most recent anyway.

A little hotter too. Was about 70 degrees (still doesn't seem that bad) with fans spinning up to 66% - And RPMs like 3500.

http://answers.microsoft.com/en-us/...gmms1sys/db3f0576-976f-4401-8793-40f01942fb9d

 

[BFG10K]

Where do I start troubleshooting this issue?

Are you overclocking anything in your system?

 

[marmasatt]

Are you overclocking anything in your system?

Nope, not at all.

I rolled back to the previous driver sent and now after about 20 minutes (a little further than usual) I rec'd the ol' IRQ_NOT_LESS_OR_EQUAL......

 

[Termie]

Crysis 3 is among the toughest game engines for graphics cards. It will crash before any other game does, but typically this is due to an unstable overclock, which you said you're not using. Is your card factory-overclocked? What model is it?

 

[marmasatt]

Crysis 3 is among the toughest game engines for graphics cards. It will crash before any other game does, but typically this is due to an unstable overclock, which you said you're not using. Is your card factory-overclocked? What model is it?

[/IMG]

 

[Termie]

OK, that's a factory-overclocked model, but the overclock is very small (up from 675MHz to 700MHz). It should definitely be stable. Just in case, you can downclock it to 675MHz using MSI Afterburner, to rule out any possibility that the core clock is causing a problem.

Another tip is to use the WHQL drivers, not the beta drivers. That means 335.23 instead of 337.50.

If you aren't having trouble in any other game, it may be that your card is on the edge of stability, and you'll need to decide whether playing Crysis 3 warrants replacing your card.

 

[marmasatt]

Thanks for the responses. I've down clocked it lower than speed (650mhz) and now I'm even getting errors on COD Black OPS and the like. I just BSOD with my 3rd different error system_service_exception.

Is my video card on the fritz?

 

[Termie]

Given that you're having problems at below stock speed and that it's happening in multiple games, I'd say yes, the card is failing. Given its age, it won't be under warranty, so you'll probably need to buy a new card. Just make sure you've tried the latest WHQL (non-beta) drivers before giving up on the card.

Something like the GTX 750 would give you similar performance for about $120. Next step up would be an R7 260X for $130 and then GTX 750 Ti for $150.

 

[3DVagabond]

While it likely won't work, you could try giving the card a good clean out. Possibly even replace TIM, if you feel comfortable doing that. The VRM on the 460 ran quite hot. If the cooler has gotten clogged it could be restricting airflow to them.

 

[FalseChristian]

I had 2 EVGA 1GB GTX 460s in SLI and they never gave me a problem. Now, another member has both of them. Your card is dying. Time to upgrade to a GTX 760 2GB (twice as fast and a much better overclocker).

 

[marmasatt]

While it likely won't work, you could try giving the card a good clean out. Possibly even replace TIM, if you feel comfortable doing that. The VRM on the 460 ran quite hot. If the cooler has gotten clogged it could be restricting airflow to them.

I actually did that and I'm still having issues. What's cleaning the TIM, the cooling fan? I was thinking about removing fan and reapplying paste but my temps simply aren't getting that high.

On a side note, I'm sorry to keep resurrecting this, but I'm reporting in as I try solutions or experience different stuff.

I've rec'd now what is my 4th or 5th error msg? Memory_management. So now I"ve gotten: IRQ_less_than_equal, page_file_not_found, some other direct X related one and now memory management. Are the different BSOD's a result of the system becoming corrupt and just causing other BSOD's or is this all related.

I also can run burn-in programs and benchmarks and it WON'T BSOD! How can that be if I'm running stress tests that try to replicate the same thing as games? It now dumps on COD and TF2. I ran chkdsk and sfc / scan now and cannot find any OS or Disc corruptions....

 

[3DVagabond]

TIM is Thermal Interface Material. As you said "paste". While the GPU could be at good temps it doesn't mean the VRM's (voltage regulation modules) are. The VRM ran quite hot on the 460 to begin with, often over 100°C typically. If airflow is restricted by the cooler being clogged with dust they could be getting hot enough to cause instability.

SOURCE: Hardware.fr

The area that is registering +101°C in the thermal image is where the VRM are located. You might have a separate heatsink over it, or it might be bare, depending on the model. If it has a heatsink over it there will be thermal tape used to interface it with the cooler. It might be dried out and not doing it's job anymore. I'm not saying this is definitely the problem, just brainstorming.

 

[marmasatt]

I can try, But that also doesn't really answer the questions of 1) Why wouldn't a benchmark/stress test reboot the system at all? 2) 4th different BSOD msg?

 

[blackened23]

Your BSODs are system_service_exception and something else IRQ related. That would seem to indicate that it isn't GPU related. Are your temps in line? I highly doubt that it's the GPU with those errors. If anything perhaps re-install crysis 3 from scratch, maybe one of the game files were corrupted somewhere along the way. It happens. I actually had Saints row 3 give me similar crash problems a couple of weeks ago for that very reason after not playing it in a very long time, re-installed it and everything was good to go.

If you have temp issues on a GTX card, you will usually crash to desktop with a directX error generated by crysis 3 itself, or you will get a driver TDR. It's very rare for a GPU issue to cause a BSOD, TDRs are usually the norm with nvidia drivers which don't hard lock your system, and it's usually because of a too high overclock. Every time i've crashed crysis 3 with a too high overclock or gpu related issue, and it's happened more than once (due to OC'ing) heh, it was a directx error generated by crysis 3. Or a TDR to desktop.


Your BSODs indicate that it isn't a GPU problem. But there's no way to be sure. Are your other games stable? Try re-installing the game.

 

[marmasatt]

No, it's not even specific anymore. I should take Crysis 3 out of the title. It's happening on TF2, COD. It just happened watching a video on a website.....But it seems to be video related. Why is it always happening watching videos or playing games. (But not even when I do benchmarks?)

 

[Anteaus]

Even mildly factory overclocked GPUs can cause stability issues and its not always related directly to heat. I have an factory OC 570 GTX that I had to throttle down to reference for some games to avoid the crashes. Now I just keep it at reference all the time.

 

[marmasatt]

Even mildly factory overclocked GPUs can cause stability issues and its not always related directly to heat. I have an factory OC 570 GTX that I had to throttle down to reference for some games to avoid the crashes. Now I just keep it at reference all the time.

Yeah it sucks, because I've done that. And this is a system I've had for a year. the only thing I've changed is installing a new SSD 2 weeks ago.

Just for kicks I used a Windows 8 disc to upgrade to see if all the BSOD's were just corruption related. But now I'm getting them in Windows 8 while I'm not even playing games. I even rec'd a new one this morning: NTFS_File_system or something like that.

Ran Windows mem test and mem was fine.

 

[Zardnok]

Yeah it sucks, because I've done that. And this is a system I've had for a year. the only thing I've changed is installing a new SSD 2 weeks ago.

Just for kicks I used a Windows 8 disc to upgrade to see if all the BSOD's were just corruption related. But now I'm getting them in Windows 8 while I'm not even playing games. I even rec'd a new one this morning: NTFS_File_system or something like that.

Ran Windows mem test and mem was fine.

NTFS File system would indicate it was a problem with your new SSD.

 

[marmasatt]

NTFS File system would indicate it was a problem with your new SSD.

I don't think so. This is like the 6th different BSOD I've rec'd now. I just ran crystal disc and it's fine.

 

[Makaveli]

Do you have a spare drive you can throw into the system and do a full OS install on?

Then test?

 

[marmasatt]

Do you have a spare drive you can throw into the system and do a full OS install on?

Then test?

Not really now, no.