• Welcome to TechPowerUp Forums, Guest! Please check out our forum guidelines for info related to our community.
  • The forums have been upgraded with support for dark mode. By default it will follow the setting on your system/browser. You may override it by scrolling to the end of the page and clicking the gears icon.

Overclocking CPU causing game instability?

What was causing this?

  • Windows 7 (Doesn't correctly utilise multicore?)

    Votes: 0 0.0%
  • It was beta. It sucka for everyone

    Votes: 0 0.0%

  • Total voters
    11
Joined
Apr 26, 2013
Messages
85 (0.02/day)
Location
Sydney, Australia
System Name Monster Gaming PC
Processor [Intel] Core i5 3570K
Motherboard [ASRock] Z77 Extreme4
Cooling [Cooler Master] Hyper 212 EVO
Memory [Corsair] 8 GB (2 X 4 GB) Vengeance Low Profile
Video Card(s) [ASUS] GTX 670 OC 2GB (GTX670-DC2OG-2GD5)
Storage [WD] Caviar Blue 1TB SATA III 7200RPM
Display(s) [Dell] UltraSharp 2407WFP (24" 1200x1920)
Case [CM Storm] Scout 2 Black
Power Supply [OCZ] ZT Series 750W
Software Windows 7 Ultimate,
Just wondering if overclocking my CPU could cause game instability. I've experienced massive, random lag spikes when playing BF: Hardline. I know the game is in beta but i can't help but feel that a CPU intensive game such as the latest battlefield releases are suffering from my CPU overclocking. I've followed this guide over at OC.net and I am getting 4.5Ghz with 1.28V. I've ran an over 4 hour prime95 small fft test without BSOD. Benchmarks show improved performance. BF: Hardline crashed occasionally. Could this instability be caused something else?
 
Bump your voltage to 1.3 and try again.
 
Bump your voltage to 1.3 and try again.
1.28 is load vcore BTW. I am using an offset. My main concern is it may not be a CPU thing as it has done multi-hpur Prime95 tests. So you are saying too little voltage causes instability just like with a GPU?
 
Yes, too little voltage can cause game instability. I had this issue when I use to run an X58 system with an i7 930 and overclocked that. Benchies and stability tests worked fine, but some games would not (I think at the time SCII was the main offender for me). Adding a little voltage helped solve this problem. It might also be worth tweaking your RAM clocks, latency and/or voltage too. All this being said though, the new BFs (4 and Hardline) I've noticed are really sensitive to any overclocking and will crash if it is not 100% stable.
 
1.28 is load vcore BTW. I am using an offset. My main concern is it may not be a CPU thing as it has done multi-hpur Prime95 tests. So you are saying too little voltage causes instability just like with a GPU?
Bump the base core to 1.3v. As long as it is under 1.4v, you're fine.
 
If it helps at all my lanbox which has a 3570k I run at 4.4ghz with 1.29v and it's all g. I could push it more but I don't have any need to tbh.
 
Just wondering if overclocking my CPU could cause game instability. I've experienced massive, random lag spikes when playing BF: Hardline. I know the game is in beta but i can't help but feel that a CPU intensive game such as the latest battlefield releases are suffering from my CPU overclocking. I've followed this guide over at OC.net and I am getting 4.5Ghz with 1.28V. I've ran an over 4 hour prime95 small fft test without BSOD. Benchmarks show improved performance. BF: Hardline crashed occasionally. Could this instability be caused something else?

Are your games unstable when the CPU is at stock settings? If not, your overclock is unstable. If so, there's another problem. Check with your CPU at stock first. Your poll is currently irrelevant and a guess for anyone to answer at this time.
 
Are your games unstable when the CPU is at stock settings? If not, your overclock is unstable. If so, there's another problem. Check with your CPU at stock first. Your poll is currently irrelevant and a guess for anyone to answer at this time.
Thanks. This problem only showed up in BF: Hardline. Not in DayZ at all. So it may be part of the game, but I started overclocking almost right before I had gotten the game. Now that the beta is over, I think I will ignore the issue until I begin to re-encounter OC caused lag. Then I will be able to test the problem's relation to my overclock.
 
beta game is beta
 
This problem only showed up in BF: Hardline. Not in DayZ at all.

So you should test BF: Hardline at stock clocks ... if you get a crash, happily reapply your OC settings.
 
i find i can have a perfectly stable oc at a set voltage. but 3dmark 11 has a better physics score if i bump the voltage up a bit. so i use that as my base line.
(find min stable voltage, then up voltage to find if 3d mark 11 physic score gets better. then find the lowest voltage i can use to get the best 3dmark 11 physics score. just look at the fps not the actuall score)
 
Yes, first try stock clock to see if crashing amount lowers .
I have seen a few reports with newer Nvidia drivers guys complaining about micro stuttering in BF4 and there fix was slight bump in Vcore (like+0.005) .
So might be worth a try but first try stock to see if it improves crashing . I run BF4 and it might crash like 1 time in a month or 2 so its pretty stable but don't know about BF-hardline , if it has all fixes from BF4 .

Also see if your getting any WHEA errors in event viewer
 
Thanks. This problem only showed up in BF: Hardline. Not in DayZ at all. So it may be part of the game, but I started overclocking almost right before I had gotten the game. Now that the beta is over, I think I will ignore the issue until I begin to re-encounter OC caused lag. Then I will be able to test the problem's relation to my overclock.

Well Dayz will not max out the GPU either as it's a flaw with the game engine has been for years now. You cannot use Arma \ Dayz as a suitable test.

Like erocker said you should test system without over clocks 1st to make sure a issue is not some thing else.
 
If you wanna test a game where the game engine is so old and bugged where multiclienting will stress your rig, try Mabinogi. That game will max a CPU after 7-8 clients.
 
My main problem here was that BF: Hardline was the only game I had which would've maxed the CPU, but then it was also the only beta. This problem is absent in all my other game so I will experiment as advised when the issue re-occurs.
 
1.28 is load vcore BTW. I am using an offset. My main concern is it may not be a CPU thing as it has done multi-hpur Prime95 tests. So you are saying too little voltage causes instability just like with a GPU?


jumping right in with a bit of wisdom i dont see mentioned anywhere else.


CPU testing programs are completely useless on their own, because a game will add a lot more power draw to your PSU, *and* up the heat in your system considerably. add on top higher RAM usage, and all sorts of instabilities can be exposed in a game, that wouldnt show up in a pure CPU stress test.
 
I use ...intel burn test/linpack , cinebench, then real game test (bf4 in my case)for stability/oc.

nothing beats that
 
I use ...intel burn test/linpack , cinebench, then real game test (bf4 in my case)for stability/oc.

nothing beats that


i use Wprime and furmark together. fast to load, and unless your drivers throttle it, stupidly high power consumption measured at the wall. if its stable there, its stable for everything.
 
i use Wprime and furmark together. fast to load, and unless your drivers throttle it, stupidly high power consumption measured at the wall. if its stable there, its stable for everything.


I was using wprime for a while and would pass yet ibt would fail oddly. I put my trust in ibt for now.

Though that is a quick test method indeed.
 
I was using wprime for a while and would pass yet ibt would fail oddly. I put my trust in ibt for now.

Though that is a quick test method indeed.


primarily for quick testing, yeah.

OCing is one of those things that is just too dynamic to rely on one test alone - you gotta test heat, max power, and then stability of each part. one game may load the CPU insanely and crash out due to ram errors, another may make your GPU overheat, and so on.
 
primarily for quick testing, yeah.

OCing is one of those things that is just too dynamic to rely on one test alone - you gotta test heat, max power, and then stability of each part. one game may load the CPU insanely and crash out due to ram errors, another may make your GPU overheat, and so on.
I think one problem with using synthetic torture apps is they generally work a very heavy but stable load, so while they are taxing the CPU, memory etc it is always pretty stable load .
Real app the load is all over the place so the voltage controller and your offset (if used) need to be good and as load come on and off . So real appp is a must , for games BF4 is "pretty" stable now, don't know about Hardline .
Realbench might be good to test, it uses bunch of real app as a BM , Handbrake, gimp ,Luxmark etc .

Wprime , how much different is that then prime95 ? , I used it few times but never as stability test .
 
Last edited:
i like to use occt with psu test. it uses a lot of gpu and cpu (alternates a bit) its designed to stress the psu out as much as possible.

also if you have the full 3d mark 11 some of the extra render tests can let you find instability when the actual bench would sail through..
 
Battlefield is very sensitive to bad overclocks. Its a good way to test an overclock actually.
 
Run it stock. Saves you a headache even with not having to reinstall windows due to a crash that corrupted the files in the os/hdd

Just wondering if overclocking my CPU could cause game instability. I've experienced massive, random lag spikes when playing BF: Hardline. I know the game is in beta but i can't help but feel that a CPU intensive game such as the latest battlefield releases are suffering from my CPU overclocking. I've followed this guide over at OC.net and I am getting 4.5Ghz with 1.28V. I've ran an over 4 hour prime95 small fft test without BSOD. Benchmarks show improved performance. BF: Hardline crashed occasionally. Could this instability be caused something else?
 
If it passes prime etc, keep it as it is and double check cpu system agent voltage - vccsa.

If you're using vccsa- auto, readjust it and set at least 0.030 - 0.050v+ offset.


btw occasional stutter can happen from wrong cpuv too, usually too low.. small 0.005v is enough.

Ie 1.230v stuttered, 1.234v was fixed. But do pay attention to vccsa, this can be crucial at times.
 
Back
Top