Computer crashes on specific games.

[Noliant]

This started randomly about 2 weeks ago,I didn't change anything,get anything new,delete anything I shouldn't have or any of that.

Now the problem is...when I play some specific games,old or new,the computer crashes,my monitor will try to switch HDMI (HDMI1 HDMI2 Analog) the sound will keep repeating (like a big bad crash),only thing that works is turning the computer off completly.

The weird thing is it's always the same games,after 10 minutes of play.
Sorry I can't give more details but...I got nothing.

I've got a SyncMaster BX2450
Processor: Intel(R) Core(TM) i5 CPU 750 @ 2.67GHz (4 CPUs), ~2.7GHz
Card name: NVIDIA GeForce GTX 285
Current Mode: 1920 x 1080 (32 bit) (60Hz)

 

[gamerk316]

Ok, basically the PC is going into a hard lock then. Something in your system went. From past experiance, hard locks tend to be memory related, but I've seen GPU's cause them as well.

If its in games and ONLY in games [or anything accelerated by the GPU, such as Flash], then the GPU becomes the primary culprit. If not, then I'd suspect the memory subsystem.

As a quick test, run memtest86 and Prime 95 [blend test]. If those both pass, then its unlikely to be memory related.

 

[Raiddinn]

How long did it work before it started having problems?

Would you characterize the games that crash as being particularly graphics intensive?

Sounds a whole lot like a PSU that is being strained to the breaking point to me.

 

[Proximon]

Are these downloaded games? Corrupt files can cause this as well.

When trying to debug these I also recommend disabling your sound device completely.

 

[Noliant]

How long did it work before it started having problems?

Would you characterize the games that crash as being particularly graphics intensive?

Sounds a whole lot like a PSU that is being strained to the breaking point to me.

Some of them are,it wasn't happening before though,that's what confusing me.
I've had this PC for bout a year now no problems,just started randomly one day.

Memtest86 and Prime 95 passed

 

[Pearlmam]

If it is the GPU like gamerk316 said, will a GPU stress test show this?

 

[Noliant]

Are these downloaded games? Corrupt files can cause this as well.

When trying to debug these I also recommend disabling your sound device completely.

Nah they're not and these games were working for a long time before this started.

 

[Raiddinn]

If you have a PC that is working for 6 months to a year and then out of the blue quits working, like 90% of the time it is the PSU.

 

[Noliant]

If you have a PC that is working for 6 months to a year and then out of the blue quits working, like 90% of the time it is the PSU.

Right but if it was the PSU wouldn't it be not working even when I'm not playing anything?

 

[Proximon]

No, power supplies are not so simple. The greater the load on the PSU the more likely it is to produce incorrect voltages or "dirty" power.

 

[Raiddinn]

It is called a Byzantine Error when something appears to work, but does something that it is not supposed to.

Say a person goes out on a warez site, downloads some brand new game and installs it and starts playing and it works. They think everything is great, it apparently worked like they wanted it to. But what if a virus was also installed?

The problem isn't the 99% of it that is working right, its the 1% that is working wrong.

In a highly complex and interconnected system like a computer is, there are very many ways that something can appear to work right when reality it isn't.

 

[Noliant]

Pretty sure I found the problem,my video card heats up to 90C/194F.
What could cause this out of the blue and how do I fix it?

 

[Keanu Reeves]

Yep, 90C will do it. What speed is the fan running at? Is it overclocked at all?

 

[Noliant]

Bout a 1000 RPM

 

[Raiddinn]

What case and PSU do you have?

Nobody should ideally want their video card to be running at 90c, but most cards can handle it if they have to.

I am still seeing this as a domino effect of a bad case leads to a hot video card which leads to a hot and overburdened PSU.

 

[Noliant]

My case is an Antec Nine Hundred Two V3 the PSU is a Corsair HX620W

 

[Raiddinn]

Those are usually pretty respectable parts.

It may really be the video card aging.

You do have a side fan and such installed in the case, right?

 

[Noliant]

I do have a side fan though I actually just took the side off and it lasted longer without shutting down.

 

[Raiddinn]

Can you aim an AC oscillating fan into it when the side is off?

 

[Noliant]

With the fan I actually got bored of playing before it shut off, then I tried the higher end games but lasted maybe 2 minutes longer than last time.

 

[Raiddinn]

So you played some low end games and it didn't shut off at all with the fan aimed into the case.

Then you played some high end games and it still shut off rapidly.

Did I understand that right?

Everything points to the video card, but my gut keeps whispering PSU to me.

The video card is clearly running at the upper limit of its range, but its just not that common in my experience for a video card that is working for long periods of time to quit working.

The thing that does commonly work for long periods of time and then stops working completely out of the blue is the PSU. - Edit - By commonly here I mean like 90% of the time.

 

[Noliant]

Yeah you got it right,I'm thinking video card since on the high end games I'll get green squares,screen flashes and sometimes a green circle that looks like the screen's burning up just before it shuts down. Is there anyway I can be sure?

 

[Proximon]

See if the card has the same problem in another computer, or try an equally power hungry card in your system.

You could monitor some temps with GPU-Z also.

 

[Raiddinn]

He already said he was getting 90c temperatures.

But yes the only way you can know for sure is to trade your PSU or video card into someone else's computer or their stuff in yours, preferably both.

 

[Noliant]

Best answer selected by Noliant.