Do I have a problem?

So, recently I've been having some weird issues with different games. Downloaded the Warlock: Master of something something demo the other day, and it would crash on startup. Worked fine after restarting my computer, but if I quit the game and restarted it, it went right back to crashing. Yesterday I bought Arkham City, and the thing crashed like Whitney Houston if I tried to get it do anything. Messing with Games for Windows Live produced a lot more crashing. Now, the thing is, I've been playing other, non-crashy games lately, new and old. And it's not like GFWL or Paradox are bastions of stability. Still, I have to wonder if the problem is on my end. All my drivers are up to date. I5 with no overclocking. 8GB Ram. Windows 7 64-bit. Radeon HD 5800. Anyone have any thoughts?

I'd start with running Memtest over night. Maybe it's an issue with your RAM, and your other games are just not using enough memory to produce the error.

If you can, verify the files.

No memory problems. All files verified. I'm wondering if it's not some kind of .net kerfuflle.

Could be a heat problem, then, or your power supply not providing enough juice. (these two things can be linked, actually.) What kind of PS do you have?

It'll probably be GPU temps that are making you crash, and I'm not sure what to use to monitor that, offhand. Anyone else have a suggestion?

So, uninstalling and reinstalling .net framework 4 seems to have gotten Batman working. When all else fails, reinstall .net.

kazooka wrote:

So, uninstalling and reinstalling .net framework 4 seems to have gotten Batman working. When all else fails, reinstall .net.

I remember reading somewhere regarding some other game that .net4 has a few wrinkles. It's the updates to .net4 that cause the upsets apparently, security updates are fine, but anything else messes it up.

Malor wrote:

Could be a heat problem, then, or your power supply not providing enough juice. (these two things can be linked, actually.) What kind of PS do you have?

It'll probably be GPU temps that are making you crash, and I'm not sure what to use to monitor that, offhand. Anyone else have a suggestion?

So, after adding an awful lot of cooling, I'm still getting these shutdowns, albeit a lot less frequently. I'm curious by what you meant in the brackets. I've got a 550 W Power supply, which is a little low, but I'm not running dual video cards or eight hard drives or anything. These shutdowns definitely seem to be associated with heat, but in my speedfan logs, I'm not getting any temps above 55, which is high, but not auto-shutdown high. (And not high for heatwave with no air conditioning.) Could be that my GPU is causing the shutdowns, but I do have a PCI fan hooked up directly underneath it.

look over the rating of the psu for the 12 volt rail and compare it to what the gpu is rated for. A few other things to try. Take out 1 stick of ram and play a bit, works? swap the other one in and remove the other? crashes? might be a bad stick. roll back/update drivers for gpu, might be something with whatever version your on. Check "event viewer" for more detail about the crash, sometimes it records a component at fault.

55 isnt high at all, dont worry about that. Are the gpu temps high? fan on gpu spinning? adding extra fans wont fix the problem regardless. How many cables run to the gpu? (6 pin, 8 pin) are they both seated properly?

Do the 5 series have fan issues? I believe the whole stuck at 30% fan speed was an AMD thing, yes?

Can you either tell us the make/model of your PSU, or else post a picture of the label on the side? 550 is usually enough, but it could be a poorly designed supply. It could also, of course, be malfunctioning.

Your first post says crashes, but the new post says shutdowns. Are you using the two words as synonyms, or is the machine actually powering off?

I'm actually kind of leaning in the direction that it's a failing card, rather than a PSU, because increasing the cooling improved the symptoms, but your total temperatures shouldn't be causing heat-related crashes unless the card were defective. This is by no means a certain call yet, but that's the direction I'm leaning -- the video card is dying.

Malor wrote:

Can you either tell us the make/model of your PSU, or else post a picture of the label on the side? 550 is usually enough, but it could be a poorly designed supply. It could also, of course, be malfunctioning.

Your first post says crashes, but the new post says shutdowns. Are you using the two words as synonyms, or is the machine actually powering off?

I'm actually kind of leaning in the direction that it's a failing card, rather than a PSU, because increasing the cooling improved the symptoms, but your total temperatures shouldn't be causing heat-related crashes unless the card were defective. This is by no means a certain call yet, but that's the direction I'm leaning -- the video card is dying.

Yeah, they're flat out shut downs, like somebody cut the power. All my fans seem to be working, and I'm at least getting motion from the fan on my GPU.

Power supply:
http://www.newegg.com/Product/Produc...

The more I look at it, I think you guys are right about the GPU being the cause. I don't think the fan on that thing is solid enough to handle the heat coming off of it in a non-air conditioned room. Don't have the video card specs handy. It came recommended from someone on this site, in some kind of daily deal website. It's a Radeon HD 5800, but the whole thing is enclosed, and there's a vent at the case exit. There's some pretty hot air coming out of that exhaust. I don't think it's failing necessarily, I think it's just getting overheated if the temperature in this room gooes north of 80 or so.

Full power failures like that are usually the power supply cutting out. That can be because the supply itself is failing, or because it's detecting a short somewhere. Typically, if you haven't changed anything in the computer to cause a short anywhere, it's the power supply itself failing.

Oh, if you have an older model AMD CPU, then it could also be a motherboard thermistor forcing a system shutdown, but I've only seen that once or twice. It's rare enough that I debated not even mentioning it, but it IS a possibility -- if you have an old AMD CPU, then there's a circuit that's designed to kill the power when the CPU gets too hot.

I lost a Corsair 650 a couple years ago, and we had someone else lose one just last month or so, so you losing this 550 isn't inconceivable.

Unfortunately, this is a pain in the ass to test. Replacing a power supply takes a long time, because you have to disconnect a zillion plugs, and then pull them out of the case, and then pull the supply, and then mount the new one, screw it down, and route all the wires back where they need to be. It's not difficult in any real sense. It won't leave you out of breath or cursing. It just takes a long time to run all the leads. And then, of course, if it doesn't work, you have to put the old supply back.

If the GPU is bad, and you replace it, then things just work. But if it's actually the power supply, a new GPU can mask the real problem, and the supply will fail later on, perhaps spectacularly. And the new GPU's changed power profile can force that spectacular failure to happen immediately, which can blow up a bunch of other stuff in your computer. I lost about $600 worth of cards when my TX650 blew.

So I'd strongly suggest replacing the PS first, because if that doesn't fix the problem, it won't do any harm. Going the other way is an easier test, but it's much less likely to solve the problem, and has the potential for catastrophic consequences. As in, every component in your computer destroyed. I've seen it happen. It just happened to me halfway two years ago -- the motherboard/CPU/RAM survived, but everything else was wrecked.

It's okay to rig up the new supply outside the case, and then just run the wires in that you need. It'll work fine sitting on a table or on the floor. That way, if it doesn't work, it'll just take a couple of minutes to detach it again.

As far as what to buy, that 550, at least on paper, is adequate for any single CPU and single video card. You made a good choice with it, and you probably don't need anything bigger. But, just to make sure, can you please tell us the exact CPU you're running, and what speed, and the exact make/model of the graphic card? And if you have an unusual number of things attached to the computer (more than 2 or 3 drives and 2 or 3 USB components), tell us that, too.

A new single-rail 550 is almost certain to be fine, this is just a sanity check.