Random game crashes, stutters, etc.

[Kyzer]

• OCCT burn test doesn't have any problems. 8m
• OCCT error detect had no problems. ~8m
• No memory errors detected by Windows, though games still output heap errors and pointer errors - exception_access_violation in particular for MC.
• Tf2 random crash and stutter, no msg
• Pd2 random crash and rare stutter, no msg
• Prtl2 random crash and stutter, no msg
• No overheating

 

• New gpu - 980ti from ebay - no artifacting and performs fine - i had stutters b4 this gpu, but crashing is new i think
• Old ram - one kit is 9 years old, 1 year old single stick, seems to have same problems
• 850w xfx ts psu - seasonic based (bronze)
• No ssd - well defragged hdd though
• Seems to have gotten slowly worse since i got a new mobo/cpu in april - asus z97-a/usb 3.1 & i7-4790k - cpu passes intel tests - i think my bios is up to date

[Nord]

First time I hear of that "OCCT" test thing, so my personal opinion here would be at first to try another CPU stresstest - Aida64 (trial) or Prime95 comes to mind.

 

Furthermore if the crashing is new since you upgraded to a more powerful GPU (as I assume?), than you did have before, it points towards the CPU.

 

CPU & GPU are equally likely to be the cause for stuttering and/or crashing programms. If you do play w/o Vsync or any other FPS cap in place, especially in the games you listed, you are not GPU bound anymore when it comes to FPS but rather CPU bound. Being CPU bound means your CPU is working at its maximum rendering the FPS for your game (and no, it does not need to be at 100% usage in that case) and hence is more likely to f* up if some other programm, in the background, requests instructions for something else. This is epecially true if the CPU is being pushed close to its limits or not knowen to run 100% stable to begin with.

Obviously a faulty OS installation or driver conflict can cause simular problems.

 

So as far as I'am concerned, either really test the flying s* out of your CPU, different stresstest, remove OC etc.

or just try a fresh windows install

or try a different CPU (which you probably wont have)

 

[Kyzer]

2 hours ago, Nord said:

First time I hear of that "OCCT" test thing, so my personal opinion here would be at first to try another CPU stresstest - Aida64 (trial) or Prime95 comes to mind.

 

Furthermore if the crashing is new since you upgraded to a more powerful GPU (as I assume?), than you did have before, it points towards the CPU.

 

CPU & GPU are equally likely to be the cause for stuttering and/or crashing programms. If you do play w/o Vsync or any other FPS cap in place, especially in the games you listed, you are not GPU bound anymore when it comes to FPS but rather CPU bound. Being CPU bound means your CPU is working at its maximum rendering the FPS for your game (and no, it does not need to be at 100% usage in that case) and hence is more likely to f* up if some other programm, in the background, requests instructions for something else. This is epecially true if the CPU is being pushed close to its limits or not knowen to run 100% stable to begin with.

Obviously a faulty OS installation or driver conflict can cause simular problems.

 

So as far as I'am concerned, either really test the flying s* out of your CPU, different stresstest, remove OC etc.

or just try a fresh windows install

or try a different CPU (which you probably wont have)

I don't have another CPU to test, Intel's quick CPU test didn't notice anything. Though, personally, I DO think it's the CPU - it could also be my RAM. I've had a lot of faulty CPU's. First 2 faulty FX-6300's, and now it's a faulty i7-4790k? Bad luck. Different case, same performance affects - the FX-6300's had booting problems and eventually died and failed to boot at all, the i7's fine otherwise - except the game stuttering. It mimics the stutter of the dying FX-6300's, but... not as bad. Mind you, I didn't have the stuttering on the FX-6300's until the night they died. I down-dated my BIOS and it seemed to clear up just a bit. Could be my board, but it appears to function fine... It does RARELY freeze when trying to go into the BIOS when it's set to EZ mode, but when it's set to Advanced Mode by default it never does. That was a long time ago, I haven't tested it, and I doubt it matters.

 

I suppose I could pick up a Celeron or something, don't know. It's most likely coincidence, but a lot of these problems with my CPU's happened after I got a NH-U12S - It's not like I'm over-tightening, I only go to where the screwdriver stops - lightly. Noctua's mounting hardware isn't anything to laugh at, either.

 

Windows has been installed a few different times on this CPU. I don't believe it's related to W10.

 

Do you know how long on average it takes Intel to do the manufacturer warranty on these things? I could wait a few days, possibly.

[Nord]

Depending on where you live, what your options to RMA are and weather it needs to be send to intel "directly" or can be tested at the shop you bought it ...  I'd say a few days up to a month.

 

I've worked in IT for a company that mainly sold computer parts and have seen basicaly everything from GPU's that literally bluescreen once you start a 3D application needing almost 3 weeks to be veryfied as "dead" by the manufacturer up to RAM sticks, that have been returned a few days after christmas, being verifyed as broken within a week.

 

 

However, if you feel like putting up with it, get MSI Afterburner and create a so called "log file" while playing (just google for it if you need a guide on how to) and post the result here. Perhaps that can give a definitive answer on which part is at fault.