Computer randomly crashes during gaming

Cravez

Hi guys

Just looking for a little bit of insight something. Recently got a new computer, good specs that are better than my old computer.

Now, sometimes and randomly during a game, the game crashes (not responding) and I usually have to restart the computer because ending task doesn't work and shutting down just idles so usually have to force shut-down.

The game(s) I play that this happens with are games that my old computer handled fine, even with the same settings (e.g Arma 2, Skyrim etc)

The new computer has the same type of liquid cooling as the old computer too.

Everything in the new computer is much better than the old one, except for the PSU. The PSU in the new computer is 750watts. The PSU in the old one was 950watts.

The Graphics Card Temperature sometimes reaches 90 degrees on these games but usually averages around 60-80, iv heard the nvidia cards can handle up to 120 degrees? The wattage on the PSU always stays relatively low (unless I am not checking correctly but the TempCore check on the PSU usually is low). I have maximum fans that my case can use.

I can't put my finger on why the computer keeps doing this mid game, everything is the same, if not better than I had previously (perhaps only the PSU being the difference) so I would like some more insight or advice into it.

I checked the hardware and it all seems to be installed correctly, but might give it another check if needed. Also my computer is not overclocked. To note this is random and not constant, I can play these games for periods without crashing, but it happens more often than it should.

My current specs:
- CASE: CoolerMaster Silencio 551 Mid-Tower Silent Gaming Case Black featured front USB 3, Internal sound damping
- CPU: Intel(R) Core™ i5-3570K Quad Core 3.40 GHz 6MB Cache LGA1155 + HD Graphics
- Cooling: Asetek 510LC Liquid Cooling system w/ 120mm Radiator (For Ultra Slient Operation -- SHARKOON Eagle SE Ultra Slient Modular Fan, 500-1200rpm
- MEMORY: 16GB (2x8GB) PC12800 DDR3/1600mhz Dual Channel Memory (Kingston HyperX Blu w/Heat Spreader)
- MOTHERBOARD: MSI Z77A-G45 Intel Z77 Chipset
- OS: Windows 7 Home Premium SP1
- POWERSUPPLY: Corsair 750 Watts CX750M Gaming Modular Power Supply
- SOUND: HIGH DEFINITION ON-BOARD 7.1 AUDIO
- VIDEO: NVIDIA GeForce GTX 670 2GB 16X PCIe 3.0 Video Card (Gigabyte Superclocked)


Cheers.

BloodBath

Those cards should not be getting anywhere near those temps. I'm guessing it's the card. How is your case airflow? Is there any dust inside the card, how long have you had it. You shouldn't really be letting current gen cards go over 75-80c especially if it's overclocked. Stability at high temps with overclocks will always be an issue. Either dial down your clocks or improve your cooling.

Cravez

The computer is only 2 weeks old, no dust inside and is clean.

The card isn't overclocked so It's getting the temps as vanilla. So I don't think I can dial down the clocking. Any ideas on improving the cooling? To note I installed windows on a fresh hard drive so would I perhaps have to overclock or change settings Asetek 510LC Liquid Cooling system?

I was reading online that 70-80 degrees was fine for a graphics card?

BloodBath

Except you said you are pushing 90c at times possibly higher when it crashes. That's way too high. It say's it's a gigabyte superclocked card so it must be pre overclocked. How many case fans do you have. I suggest adding more. You want at least 2 intakes and 2 exhausts ideally.

Cravez

Can you dial down the pre-clocked cards?

I have the max amount of fans in the case I can use and cannot add any more. I would have to check if they are intake or exhaust (can't at the moment) but I think it should be as you described.

BloodBath

Well there is no way you should be getting those kind of temps then.

Can you link to a card the same as the one you have? You said it's a gigabyte superclocked but I think only evga uses that term for it's overclocked cards. Gigabyte call their's OC's.

That standard evga cooler is pretty crap which might explain the temps. They still shouldn't be that high though imo.

You could also try it with different sticks of memory. Try it with 1 stick at a time to check if it might be a faulty memory module.

uncle_sam_ie

My MSI GeForce GTX 670 OC gets to 78c. But, this is normal for the card. 90c for any card is bad though.

Cravez

I'll have to check the card when I'm home, the only description I have at the moment is very generic.

I'v heard MSI Afterburner can work well on increasing fans peeds when GPU's reach high temps. I had it installed on the old computer but never used it, the old computer had manual adjustments on the fan speeds.

jimmycrackcorm

Michael Cravez wrote: »

Can you dial down the pre-clocked cards?

I have the max amount of fans in the case I can use and cannot add any more. I would have to check if they are intake or exhaust (can't at the moment) but I think it should be as you described.

Did you add them yourself and if so - are they all pointed in the correct direction?

BloodBath

Yes you can make custom fan profiles with it and adjust clock speeds and voltage. You might be able to undervolt it a bit and dial up the fan profile a bit to get temps down.

Cravez

btw it doesn't constantly reach 90c when I play the games. It averages as I said about 60-80. I'v only ever seen it reach 90 about once or twice.

Cravez

jimmycrackcorm wrote: »

Did you add them yourself and if so - are they all pointed in the correct direction?

Nope the computer was pre-installed. I can't say for sure if they are in correct direction (the same crowd that did the old computer which I had no problems with). I'm still a bit n00bish with these things

dyer

70 degrees at full load sounds reasonable to me.. you might push higher depending on the environment (central heating on etc) .. you could try increasing your fan speed via nvidias control panel and see if that helps. the front fans should be configured as intakes and the rear as exhaust.

have you tried stress testing your rig or running diagnostics on your ram modules?

grab a copy of superpi and let it run on all cores.. also wouldn't hurt to get memtest86 and checking your ram. i'd go with checking the ram first.

BloodBath

The side the sticker/label is on is the direction the air is going.

Tails142

The nvidia gtx 560m card in my laptop reports temps of 90+ when gaming and i have used it no prob for over a year so i wouldn't immediately say temps are your problem.

Heat can cause the problems you are having though so is there space around the cards inside the case, can you move cables away? Is there anything close by the exhaust area of the case have you got the case under a desk or close to a wall where the heat is getting trapped? Is the ram getting cooled by a case fan, is the north bridge or other motherboard fans working right and pointing in the right directions?

Are you using the latest video drivers?

Cravez

Thanks for the advice so far, ill check things out when I'm back home later and update.

Cravez

Tails142 wrote: »

The nvidia gtx 560m card in my laptop reports temps of 90+ when gaming and i have used it no prob for over a year so i wouldn't immediately say temps are your problem.

Heat can cause the problems you are having though so is there space around the cards inside the case, can you move cables away? Is there anything close by the exhaust area of the case have you got the case under a desk or close to a wall where the heat is getting trapped? Is the ram getting cooled by a case fan, is the north bridge or other motherboard fans working right and pointing in the right directions?

Are you using the latest video drivers?

The cable management in the case looked fine. Although I will have another look later. The Wireless card was close to the GPU off the top of my head. The computer is under a desk but away from the wall. I can check the others later.

I have updated all drivers for my computer.

dyer

seeing as it's a new build, check if there's a new bios update available for the mobo while you're at it.

BloodBath

The nvidia gtx 560m card in my laptop reports temps of 90+ when gaming and i have used it no prob for over a year so i wouldn't immediately say temps are your problem.

The 500 series was notoriously hot. His card is overclocked although not highly overclocked. I know from experience that stability becomes a big issue with heat that would normally be ok the higher you clock the card. If he's pushing into the 90's this will become a problem with even small factory overclocks.

Cravez

dyer wrote: »

seeing as it's a new build, check if there's a new bios update available for the mobo while you're at it.

BIOS is already updated.

dyer wrote: »

The 500 series was notoriously hot. His card is overclocked although not highly overclocked. I know from experience that stability becomes a big issue with heat that would normally be ok the higher you clock the card. If he's pushing into the 90's this will become a problem with even small factory overclocks.

If this does become an overclocking issue, can I dial down the card even though it may have been pre-clocked?

danger_mouse_tm

Any chance you could test with one stick of someone elses ram? It solved an ongoing issue for me. We discovered that we had to change out the ram completely on an older MSI board to stop it crashing.

Cravez

danger_mouse_tm wrote: »

Any chance you could test with one stick of someone elses ram? It solved an ongoing issue for me. We discovered that we had to change out the ram completely on an older MSI board to stop it crashing.

If every option is exhausted I will try this.

Overheal

What is your CPU Temp?

I have Crossfire 5770s that sit at 40 deg and a CPU that sits at 19. Both are air-cooled. Your issue sounds like cooling, frankly.

You should ensure your case is a Positive Pressure Airflow environment. For one, you should always have more fans blowing into the case than blowing out. If too may fans are blowing out, you're going to get dust issues later in really bad places (everywhere). If too many fans are blowing in, your heat has few ways to escape.

Snap some pics of your case and lets see what youve got.



You can see in this photo, theres a 200mm fan that shoves colder air down into the case, 1 exhaust case fan ~120mm, which is tandem with (wow old photo) a newer CPU cooler that pushes air parallel to that exhaust from the CPU. Then 2 GPUs handle their own air intake and exhaust, which they each get from 2 120mm case fans that sit in the front of the build. The PSU at the bottom of the case has its own 120mm fan that helps pull air down from the top 200mm fan, the front 120mm fans, and shoves all that hot PSU exhaust out of the build before it can contaminate the air flow at all. In all, thats the CPU fan, a 200mm intake, 2 120mm intakes, 2 GPU and 1 PSU dual-action fans (manage their own airflow basically), and 1 120mm exhaust fan.



Airflow is basically two big channels pushed front to back by the front case fans that each get used by the GPUs and PSU/Hard Drives, then the CPU gets its cooling from the 200mm top fan, its cooler, and the rear exhaust. I could probably increase the speed of the rear exhaust from low to medium and get even better numbers; right now the 200mm fan has a bit of pushback (paper will hover on top of it rather than cling on)

Reassess how you've configured your airflow and ask yourself if you've created any pockets of air that might be stagnant.

Cravez

I think my CPU temp is fine.

I'll get some photos up later. What i'll do is take a snap of the temps when running one of the games. I'll take a few snaps of the computer case as well

I'd say I just don't have any configuration on how to manage the cooling, I was able to manually adjust it on my old computer (always kept it at a medium) so it's probably what I need to do here also.

BloodBath

This is basically how you want your airflow depending on what case you're using.

yoyo

Check your task manager for a process called "iehighutil.exe", this is a bitcoin miner malware, which can get installed when installing game mods etc. from dodgy sources. This will cause your GPU to max out and possibly crash it causing such instability. It should be removed if you have it, kill the process and look for it in the C:\temp iirc. Had to remove a couple of these infections, av software generally doesn't pick it up!

Nick

Cravez

Ok guys got a few pics. When I checked my case, I have id say from the looks of it about 1 exhaust and a few input fans. I don't have a fan or exhaust on the roof of the comp (my old computer case did). I still couldn't really tell much about the fans

First one is of the temp of the CPU's and the GPU when playing Arma 2: http://postimg.org/image/xjbvl0dc1/full/

Here's the screenshot of the overall computer inside: http://postimg.org/image/3lbi60daj/

One of the CPU + Cooler Fan: http://postimg.org/image/9xvah9v2j/

The Graphics Card up close: http://postimg.org/image/ojy2dduu1/

@Bloodbath: This looks to be the graphics card in question

Shot of the front fans: http://postimg.org/image/6tld18is7/

Power Supply: http://postimg.org/image/a41b1nckj/

Front Fans infront of comp without tray: http://postimg.org/image/5gwrqfzg9/

Fan Under the PSU: http://postimg.org/image/9avicf2il/

Behind computer shot: http://postimg.org/image/hwytkf17l/

I know it's alot of pics but just wanted to try get as much as I could. Let me know if anything sticks out or needs more pics. I might try MSI Afterburner now.

Cravez

BloodBath wrote: »

Yes you can make custom fan profiles with it and adjust clock speeds and voltage. You might be able to undervolt it a bit and dial up the fan profile a bit to get temps down.

Any tutorials online for MSI Afteburner that you know of?

dyer

MEMORY: 16GB (2x8GB) PC12800 DDR3/1600mhz Dual Channel Memory (Kingston HyperX Blu w/Heat Spreader)

is the ram configured correctly in the bios?

when you're all booted up, you can use cpu-z to check the timings.. it will also show you the spd values at various voltages/clocks.

if all that's straight, next thing i'd do is remove a stick and see if you still get freezes. if you do, try the other stick.

the ram would be the easiest thing to rule out at the beginning.. and it's often the culprit with new builds. sometimes a stick of ram just wont sit pretty in one slot but will work fine in another.

Overheal

on newegg owners talk about the card idlingat 25-30 and peaking at 50-65 under load. One nutf*ck with 3 in Triple SLI has them peaking at 69.

In your case there looks to be an awful lot of obstruction that the 2 front case fans have to overcome. the case itself has its mounts set up to be natural airflow deterrents, and your hard disks I would separate out a little bit personally, or at least mount them to the bottom of the chassis to give you a lower center of gravity.

If you can install a fan on your side panel, I'd strongly recommend it. For now though try running it with the case open and/or the slot-covers off.

BloodBath

Looking at the case I think airflow is the problem. There is no bottom intake. The front intakes have to push through a vent and then through the hard drive cages. Do you leave the door closed while gaming? Is there side vents on the door frame so it can still take air in if it's closed?

There's then only 1 exhaust which is also acting as the cpu cooler.

I would suggest replacing that case at some stage or modifying it.

That model of graphics card has one of the best coolers. At stock speeds you really shouldn't be going over 60-65c. It's not 100% the cause of the crashing but it is an issue.

Alt tabbing out to check temps during gaming is not ideal as the core temp will drop straight away. You want something that will record your max temps while playing. I think afterburner does this as well.

The fan profile is easy to set up. It's as simple as clicking on the graphic and setting the curve the way you want it.

For example here is the way I have mine set in Trixxx which is similar to Afterburner but for ati cards only.



I only have 3 points set. It slowly ramps up until it get's to 70c then ramps more steeply up to 100% at 100c. It never goes above 70c though.