Frame stuttering, need help!

jakewat97

Distinguished
Dec 15, 2012
206
0
18,710
As the title says, I am experiencing often frame stuttering after X amount of time gaming where the frames will begin to stutter/ become choppy and the GPU load visually displays very slight sometimes large spiking on gpu-z. I have been able to rule out the graphics card as I have tested a 560ti, 770, and 760 with nothing changing. My motherboard, CPU, Ram and power supply are only a few weeks old and I had my 770 sent in to be examined before I had tested the 760 as originally I thought it was the cause. My only real idea now is that the CPU is not being cooled sufficiently by the stock cooler (waiting on my W/C stuff) but am not 100% sure.

Extra info: graphics card (760 in this case) never exceeds 60, cpu gets to around 67-68 at full load which is not on all games, motherboard has 3x 4pin cpu power connectors which I have all connected to make sure of sufficient power (even though the extra 4 is for OC), Often a restart will fix the problem for and hour or so, driver rollbacks and wipes and clean installs haven't helped.
Specs:
Z87X-OC
GTX 760 (atm)
8gb kingston hyperX beast
850w strider gold PSU
TJ07 case so sufficient airflow
stock cooler

Any help would be appreciated as I have had no luck with past forum posts on this problem.
 
Solution
@woltej1 I wouldn't think ram is the problem, just installed some that had been previously tested on another pc to be fine and they are only about 2 days old and have been loaded with optimized defaults. Even with my old ram (hyperX blu) memtest showed clear. And yes driver rollback meant I installed 2 different older drivers with no luck. And it happens on all games, but usually after an hour or so of playing so it is very annoying to test and replicate.

Two things I have lying around to test are: an ssd with windows 8 clean install for testing with a single game to cut out the possibility of a software issue, and also a noctua cooler to test for cpu temp problems. which I will hopefully set-up today.

Also today I tested gaming again...
Is this all games or just some? I had a similar problem my gtx460, can't remember what game it was, but installing an older driver solved it, is that you did by "driver rollbacks"? Also reinstall the game if it's an isolated game. If it's after a certain amount of time I'd think temp, but you already checked that. Could also be your RAM. Have you double checked the timing and voltages in your BIOS to make sure they're right? Also run a MEMtest. Once the RAM heats up it could start leaking.
 

jakewat97

Distinguished
Dec 15, 2012
206
0
18,710
@woltej1 I wouldn't think ram is the problem, just installed some that had been previously tested on another pc to be fine and they are only about 2 days old and have been loaded with optimized defaults. Even with my old ram (hyperX blu) memtest showed clear. And yes driver rollback meant I installed 2 different older drivers with no luck. And it happens on all games, but usually after an hour or so of playing so it is very annoying to test and replicate.

Two things I have lying around to test are: an ssd with windows 8 clean install for testing with a single game to cut out the possibility of a software issue, and also a noctua cooler to test for cpu temp problems. which I will hopefully set-up today.

Also today I tested gaming again and found the cpu load is also spiking on task manager.
 
Solution
Tested as in they were installed and it booted fine or was tested with a program. As time goes on the system will cache more and more on your ram so as it fills up it may start to hit areas that leak. I wouldn't expect cpu to be an issue as you wouldn't see throttling until a handful of degrees before the max temp. If info that is suppose to be cached leaks, your cpu will spike to get it. It's worth a test, just boot the CD and go make yourself something to eat and it will be done by the time your back.
 

SlaKer440

Honorable
Apr 12, 2013
58
0
10,660
Your GPU temps are fine. 60degrees is fine (dont go above 80C) however your CPU temps seem a bit high. I believe the recommended Intel spec temps are 72C. Throw on the Noctua heatsink and see if it helps. That is my only guess as your GPU is well under recommended max temps (90C)
 

jakewat97

Distinguished
Dec 15, 2012
206
0
18,710
They haven't been tested yet. Windows 8.1 is already installed on the ssd so I won't need to wait, I just hate the thought of using that piece of OS nightmare. I still have no idea how the ram would be causing the problem though as the system itself is practically new with a barley month old installation of windows 7, and on my previous system with the same old ram, 2500k, 560ti, p67 etc I never had an issue. The only two things left to me really are the cpu or the motherboard.

keep leaving ideas guys, I'm off to test these other parts.
 

jakewat97

Distinguished
Dec 15, 2012
206
0
18,710
They haven't been tested yet. Windows 8.1 is already installed on the ssd so I won't need to wait, I just hate the thought of using that piece of OS nightmare. I still have no idea how the ram would be causing the problem though as the system itself is practically new with a barley month old installation of windows 7, and on my previous system with the same old ram, 2500k, 560ti, p67 etc I never had an issue. The only two things left to me really are the cpu or the motherboard.

keep leaving ideas guys, I'm off to test these other parts.
 


It sounds like you've been having this problem for a while and have done some testing, but not on the RAM so it seems at least worth it. and you said that you didn't have this problem with your old RAM? that's something pointing to the new ram being an issue.
 


I wouldn't say it's hot enough to cause problems. My OC'd i5-750 peaks at about 60 with a CM hyper 212. The night before it arrived I played BF4 on my stock cooler so I knew if the new one would make a difference and I peaked at 95C. I didn't notice it while playing, just when I checked Realtemp afterwards.
 

SlaKer440

Honorable
Apr 12, 2013
58
0
10,660


Yeh, it doesnt seem hot enough to cause problems to me either however I know of intel CPU's that drop performance when hitting mid to high 60C most likley due to a bad batch. CPU's are a bit of a lotorry when it comes to OCing so you never know.

 

jakewat97

Distinguished
Dec 15, 2012
206
0
18,710
Sorry I'm not explaining my ram situation very well so I'll go step by step.
old system with hyperx blu ram: no problem
new system with hyperx blu ram: problem
new system with hyperx beast ram: still problem
New system hyperx blu/ beast with optimized defaults: still problem
That is why I find the ram to not be the problem.
I just installed the noctua and will begin testing soon. Just a note, the new haswell cpu's are worse than say woltej1's i5-750 in terms of heat as the thermal compound used between the chip and board is some cheap s**t, it also states on intels site that my 4670k has a TDP of 75 degrees +/- a few degrees, that is why I am suspicious of a heat issue as I am getting close (67-68ish degrees)
 

jakewat97

Distinguished
Dec 15, 2012
206
0
18,710
Yes, no matter which I use the problem persists. Although in the bios I did notice the timings where at something like 11-11-11-27 which is wrong, I may go back and load the xmp profile of 9-9-9 which is the correct timings
 

jakewat97

Distinguished
Dec 15, 2012
206
0
18,710

usually hours, as in 1-2, but sometimes it can happen in half an hour or so.
 

jakewat97

Distinguished
Dec 15, 2012
206
0
18,710
Ok so for anyone that is still following I think I fixed the issue but will continue to watch out for anything. Only problem is, if I have fixed it I do not know, out of the three fixes I tried, which worked: the cpu cooler, windows 8.1 fresh install, or memory timings corrected.

But anyway, thanks for the input guys and hope this may help anyone if they ever have the same issue.