It's just crysis 1 that doesn't like my OC? WTF?

so after going with stock clocks, it rebooted my pc in about 1 minute gameplay. Something else is wrong and it's not the overclock....
Before i managed 45 to 50 minutes of gameplay with the OC.
Going to test the memory. If that passes i'm betting my Money on my 7 years old PSU...

 

Yep sounds like you need a new one.

 

Same for me with Crysis, had to back mine down to 1070/1500, as when testing OC's with the bench, on the 2nd run through, id start to get yellow/orange polys, yet everything else runs perfectly fine with the sliders in the CCC fully maxed out.

 

Yeah i think that game is still very demanding even for today's graphic cards.
But mine is simply rebooting now even with stock clocks. when it's the first time i play in the day, i can generally play 45 to 50 min. Then it reboots within a minute of gameplay.
Tested 14 hours of memtes86+ yesterday, no errors reported.

 

^Def sounds like PSU, you should get it checked. Did u not replace it?

 

Nope i'd sticked with it. What i don't understand is the psu is stable when running bf3 for hours. Why only crysis?
Lost my confidence on this Psu, going to replace it for sure ASAP

 

Oh I misunderstood you, I thought your rig was rebooting with every game. It might not PSU then...

Have you tried playing with everything at stock clocks?

 

Yes. the game reboots the same. Normally the first time it takes longer to reboot. Yesterday left the computer running memtes86 no errors for 14h.
The cpu is new, my old one was rma to intel 2 mouths ago when it suddenly died. I opened a post 2 mouths ago when that happened. Dunno if the 2 events could be related
"http://forums.guru3d.com/showthread.php?t=378835"

 

BF3 MP or SP? Any other games crashing? And at what settings?

Can u list what test you've run, like OCCT, memtest, P95 or 3DMark 13 for example, did they all pass?

 

I play a lot o bf3 MP and it's fine. No other games have done this. Not even crysis 3. Just crysis 1.
I've tested my cpu overclock with prime 95 and intel burn-in test for 8 hours. Temps are all fine. Runned memtest86 yesterday no errors. tested gpu overclock with uniengine valley and heaven.
Never runned 3d mark 13 or OCCT.

 

See how 3DMark 13 runs.

 

Downloading it now. Let you know how it goes.

 

3d mark 13 ran ok no problems. Max temps on cpu were 55cº. GPU 65cº. And i ran with both the OC on the cpu and graphics.

 

PC rebooted now after 2 hours of planetside 2. It's not just crysis and it's getting worse. Could the psu be overheating? she is clean, nothing is obstructing her (she's at the bottom of the case). Gonna try to remove the dust filter

 

In my experience, the pc suddenly rebooting is usually a sign of something being off with the PSU indeed.

I guess it's time for a replacement, hope it works out and the problems go away!

If a replacement doesn't fix it, then at least you know for sure it wasn't the PSU

 

Yeah you should prob replace PSU, it is 6 years old after all.....

 

Problem solved guy's! It turns out that my old PSU was overheating, shutting herself down. I tought that all the fan noise i was hearing was from the HIS 7970 iceQx2 that i bought, but turns out it was the psu running it's fan to the maximum speed. And here i was, telling my friends the IceQ x2 cooler was very loud lol. Pc is super silent now running games with my brand new Corsair HX 750w. Been playing crysis no probs for 3 hours, no reboot or anything.
How the PSU started overheating all of a sudden is behind me, but what the hell, all is fine now. Thx's for all your help gurus!

 

Curious question (will tell you why later), but was the "reboot" that was happening a shutdown then a bootup?

 

The pc never shutted down completely. It behaved exactly as if i pressed the resset button on the case.

 

100% sure.

I set my clocks to 1200/1650 @ 1.2v and in GRID2 which was the best way for me to test as when you view your car you can spin around it and I would get the same flashing black/white chequered squares in the walls behind the car.

It wasn't random either it was always the same place, but sometimes it was a big square and other times it was a small patch. So I lowered my memory clocks to 1600 and still got them, then 1550 still got them, then I went back to stock memory clocks and left my GPU core at 1200 and low and behold the squares were still there.

So I rammed up my memory clocks back to 1650 and lowered my Core clock to 1150 and BAM no squares at all in GRID2 I even sat there for about ten minutes just spinning my car around to check for the error and nothing showed up.

So I test Metro LL for a bit. I got the same thing in that game as well before, I played for a bout an hour and nothing at all appeared in the places I knew it always happened. So I went and tried to get better temps as my core voltage was pretty high at 1.2v for 24/7 so I lowered it back to stock 1.133v (that's what afterburner reports as stock), which I know these clocks (1150/1650) are NOT stable at and I slowly raised the voltage on the core until I lasted for 30 mins in Kombuster I got to 1.138 which is a VERY minor increase and it lasted for 45 mins with no higher than 58C on the core and VRM's were 67C each.

My memory voltage is stock 1.6v and so far I have had no crashes in any game at all, and those black/white squares have vanished. I think I have a really good chip my ASCI qualitu is 78.2% so its pretty good. I also changed the thermal paste on the card for some IC Diamond and it helps a lot. Idle temps are now 28-30c and load is 55C after 2 hours of BF3.