Hi there First Specs Corsair AX850 http://www.corsair.com/en/professio...lus-gold-certified-fully-modular-power-supply 4790k @ 4.6GHz @ 1.28V Corsair h100i Asus rog vii ranger (latest drivers en bios) Windows 10 (latest updates) Galax GTX 980 TI HOF (latest drivers) ADATA XPG 16GB DDR3 2400MHz RAM 2x SSD (128GB + 256GB) Corsair Air 540 case PG279Q My issue is that when I play crysis 3 on not all maxed graphics but sort of max graphics my pc will reboot. No BSOD nothing just a slight freeze then reboots. Now been searching online about this issue and by the looks of it, it may be the PSU that's not stable under super heavy loads on the 12V rail. I find it hard to believe that my psu is not enough but maybe it's on it's way out (maybe not soon but it's not stable anymore). I've played the following games with no freeze/restart so far. -BF4 (ultra graphics minus post AA and MSAA played for about 1-2hr's straight on max graphics) -Doom (played it Wednesday for almost 2-3hr straigt had 1 crash but i think it was game releated as I had to teamviewer from another pc to close the game but without even rebooting game was running fine) -Project cars (tried twice and did a few laps in the rain having my gpu work 100% for about 99% of the time also driving in sunny weather. Have only tested it for about 30min max but had it standing on a spot where it uses my system the most and it worked for about 40min before I decided to quit) -Ashes of the singularity (Doing the cpu and gpu benchmark and it works 100%) -Cinebench a few times and it always completes -Nvidia Pendilum Demo (Trust me that thing stresses out the gpu a lot. My power usage on my 980 ti even on crysis is not above 100% but on that almost always near or above 100%) -Aida64 for about 20min with no crashes (stressing memory together with cpu and FPU) -Heaven benchmark Now my CPU temps is max about 70C-75C when playing Crysis 3 in extreme cases but doesn't stay there for very long then drops to mid 60's and in Aida64 it's about 78C max. i have a custom fan curve set in MSI afterburner and my max GPU temps is normally about 71-72C but Crysis is the only game that can get my gpu up to 75C+ but then again the Nvidia pendilum demo also stresses out my card a lot and get it near or past crysis level of temps. Maybe because the Pendilum demo only really uses my gpu the PSU can still handle it but in Crysis3 it uses my cpu and gpu to much and then my psu can't handle it anymore. Not that my PSU is not enough their might be a fault with it. So I want to know how do I figure out what is the cause. Like I said ran aida64, Heaven and so on but maybe not for long enough periods of time. Maybe someone can give me a tool that will stress test the psu like have my gpu,cpu,HDD and pretty much everything running at the same time so that my PC can almost get to it's absolute maximum power usage it can physically draw from the psu to see maybe my psu is not stable under heavy loads. I was at a friends place and when we compared crysis3 on both systems my pc had a BSOD related to not enough Vcore so maybe my CPU is still not stable. I know I should try setting my overclock settings to default and test again but first is there a way to test my PSU so that I can do that then see if it triggers the issue and know how to test it when i reset my bios to defaults? Hope this post is not to long but I just want to get as much info out there so that you guys can better understand the issue here
Do the event logs say anything? Windows Logs > System Look for events with ! Error Critical just before the time of the reboots
No nothing about that. All i see is kerner power error. the thing is my pc dosn't switch off the on. i think it should switch off when the OCP kicks in but it's just like you press the restart button. And the weirdest thing is that I don't even see any BSOD. I little while ago i was at my friends place and I got the BSOD on Crysis 3 that was related to CPU voltage. but how come my cpu is stable then in Aida64 having everything ticked except for GPU test and local disk test.
Just watched a video about the RX 480 power draw issue, and the guy was saying generally the system just powering off like that is either overheating CPU, which you have ruled out, or the PSU not giving enough power If you were having CPU voltage issues, along with this problem that could be PSU related, I think I would test a new PSU Could be that the game is stressing the CPU in a different way that causes the problem, weirdly, I had my old Phenom II Prime95 stable, but when I played the original crysis, it bluescreened I'd test a new PSU, if for nothing else than to rule it out
I recently had a similar problem after installing my H115i cooler. I had overtighten the screws which caused the computer to freeze. Turned the screws on the cooler 1/4 back and been running all fine since then
Did your pc freeze and then reboot or did it first go completely off as if someone is pulling out the plug please describe it in detail what happens. So far the games that works for extended period of time like at least 30min-1hr+ is Black ops 2 (I know not very demanding but include it) Doom (1-2hr) Project cars BF4(I play it pretty much every night) Rainbow six siege (on my graphics settings it's not that demanding have yet to test on my PG279q coming from my Qnix qx2710 @ 96hz) GTAV So far it's only Crysis 3 giving the issue. not freezing on desktop or pretty much anywhere else for that matter. Reading up a bit seems like that game can do that to a system even at stock clocks. I also don't have another psu to test.
That's what I also thinking unless something is broken like a faulty cpu but even that might not do this. From past experience you will have either get BSOD and then reboot, pc will just freeze and stay there or pc will shut off and you have to either unplug and plug back in or just switch on again. Nothing like this and seeing it's only Crysis 3 maybe my particular system and this game giving me issues. Do you people think it could be a "hidden"feature in the motherboard where it detects the freeze and then reboots automatically. Also hard to believe it's ram as i think my pc uses more ram when playing BF4 than this game. I had my gpu overclocked to 1430mhz Core 3703mhz Memory for about 3-4months and then i was testing out COD:Ghosts and it was running ok for 1hr and then freeze. Made my overclock so unstable it doesn't even like overclocking anymore. I can now only do stock memory and 1420mhz Core clock so maybe even my gpu is not as stable as I though it was. but then again doom also stresses your GPU a lot or so i think and was playing just fine for 1-2hr maybe 3hr didn't take note how long I played.
Have you got any other hardware you can disconnect from the PSU to lower the load? Or drop your CPU / GPU OCs for a while to test Give the PSU as little to do as possible and play C3 again, see if it still does it Board would shut off the machine if the CPU was overheating, so I guess it's possible the board is doing it, but I'd lean towards the PSU first, they are generally the cause of random weird problems like this, and generally the last thing people think are the problem
I understand what you saying but what I just can't understand if it's the psu to my understanding it will rather switch off than to just restart the pc without it switching off. I can lower the clock and then tet gpu and cpu at once on the load. I was also thinking of running heaven or furmark and aida64 at same time
It just completely froze. I had to do a hard reboot, which why I said "similar" problem It froze after idling for several hours, but worked perfectly fine under extensive gaming sessions and other work.
Let me explain again. I play crysis 3 for exmaple 30min then pc freeze and restarts without it going off completely or even remotely. I hae that option disabled in the bios. Mine almost the same just imagine you not having to press the reset button it happens automatically. PC doesn't go off completely it just freeze and immediately restarts almost like you pressing the reset button on the pc. Like i said it only happens on crysis 3 for now.
Yes, I understand what is happening, but you seem adamant it is not PSU related, I'm showing how the symptoms could still be PSU related, PSU cuts out, BIOS powers the machine back on thinking there was a power loss, to you, it looks like the PC was reset and it just reboots I guess even with that setting disabled, if the cut out was short enough, it would be like you hit reset, seen it happen to mine when there was a real power cut for a split second, enough juice left in the PSU to not 100% lose power, but enough to reset the machine It might not be the PSU, but don't rule it out is all I'm saying
This kinda makes sense to me. if the power cut is so short it's like it's never off. A ups is a great example. When you have a inline ups you do get a few milliseconds o power cut before it switches to battery thus the pc will not go off.
Yea exactly, I've had a power cut so short that the TV went onto standby, the lights went off and back on, house alarms in the street went off, but the PC kept running like nothing happened, other times it has been just long enough to reset the machine And other times when it went off, on, off, on, off, on > me run to the PSU and disconnect it myself before something exploded
When my psu died it keps buzzing for 3 seconds in WoW and then just randomly shut off. Does your pc instantly boot back up or does it have like a period where its just dead for like 5 seconds?
I've had a PSU cause BSODs for weeks and weeks, had absolutely no idea what was causing them, reinstalled Windows, did everything possible to cure it, could not work it out until it went bang and stank the room out with burning RMA'd the PSU, got the replacement, no more BSODs They can cause a whole host of strange problems that you'd never ever consider the PSU being at fault for
Not true. Unstable memory can cause this, mainly in my testing low system agent voltage will cause reboots. Of course this could also be any voltage for memory subsystem, dram voltage, vttdr, vccio etc. Another culprit can be low input voltage. @OP 20 mins of aida64 is not enough. Run memtest pro to 400% and p95 blend(older version that doesn't have avx)
