Asus AI Suite 3 - Getting a BSOD while CPU temperature is around 60C

unjoyer

Reputable
Sep 4, 2015
5
0
4,510
Hi guys,

I'm using AI Suite 3 to OC my CPU. Yesterday I bought a NZXT S340 case and Corsair 100i GTX water cooling. Getting nice and chill idle temps for my processor with this.

Anyway in AI Suite 3 I did the extreme tuning and got to 4700 Mhz while monitoring my temperatures which were nice around 64C and got a BSOD. Sadly I didn't catch the 0x000xxx from the BSOD, but I will if neccessary.

Someone is telling me that I can still up my clock by messing around with voltages because temperature wise I still have room to tweak, but I have NO idea how to do this in AI Suite.
Let it be known that I am not too comfortable nor competent to go and do this in BIOS unless someone walks me through it. (So please keep your "baa-haa OC'ing with software" stuff to yourself. It's not much different as the software tells the BIOS what to do instead of me in the actual BIOS)

So if you guys think I could up my game with my current set-up I would appreciate any know-how on how to do this.

I've added a screenie of what's currently happening in my AI Suite 3

Specs:
OS: Win 10 64bit
CPU: i5 4670K Haswell 22nm
MOBO: Asus Z87-A (1150)
RAM: Kingston 16GB DDR3 2400Mhz (KHX2400C11D3/8GX)
GPU: Geforce GTX 970 Gigabyte Windforce
Case: S340
Cooling: Corsair 100i GTX
PSU: Corsair CX500


RgUXiHj.jpg
 
Ive got the I7 version of your chip I just went through all of this last week. According to your screen shot you are already over volted. Your chips stock voltage is around 1.135 you are running at 1.275 (you may be able to go as high as 1.4v).


You should you Intel Burn Test to test your overclock (its a free download, just do 10 passes on very high, watch your temperatures be ready to stop if you hit 90°). When your testing you overclock switch from adaptive mode to manual mode. (very important-adaptive and stress tests dont mix).
Try running 1 pass of Intel Burn on the settings of your screen shot to make sure you really are stable.

If youre stable at 45 (based on the screen shot) try bumping up the OC voltage to 1.285 and the multiplier to 47 and run intel burn test see if you blue screen or get too hot. If everythign is ok increase your ratio , run intel burn test, keep increasing your ratio until you get too hot or blue screen, then increase you voltage by 0.005 until you find the magic number.

Intel Burn Test will get abouut 15° hotter than what you can really expect as long as you dont pass the 85° or so youre ok.

Change back to aaptive mode keeping the same voltage and multiplier when your done stress testing.
 



That's exactly the kind of attention I was looking to get from this site. Thanks so much I will do this first thing in the morning lol and post back my results. How long would you suggest me running the stress?
 



I think 25 passes on very high is great/maybe overkill (do that when youre pretty certain you've found the right muliplier voltage ) - When your just testing to narrow down your voltage/multiplier 2 or 3 passes will give you a pretty good idea if you are stable.
*Intel Burn Test if pretty intensive. All stress test programs work a little differently, Intel Burn lets you know if your stable the fastest but also gives you temps higher than what you'll see in intensive gaming

To verify Intel Burn Test' download AIDA64 and run that stress test for 6 to 12 hours.

Info you dont need: Running Prime95 small FFT test for 12 hours is what most people use to confirm stability, BUT Prime95 isnt compatible with Haswell chips (like yours and mine)