[Solved] CPU Overclock is being a pain, help..

[John93]

Hello!

I will get into the problem straight ahead. So i have an Intel i7 3930k CPU and i have been running it at 4.7GHz for about 5 months now, and i haven't had a problem with voltages or anything right untill recently, it started BSODing on me whenever i'm playing a game, this happens about 15 minutes into the game. when i first OC'd, i could run the CPU at 4.7 at 1.405v & i remember that i tested with prime95 and it went all good, but now, it seems like it's needing even more voltage suddenly ?! i tested all possible OC settings but couldnt get any stable below 1.430v which is quite high compared to when i first OC'd back in february. I'm using the same components as i did in february, the only thing that i did recently was custom water cooling loop, and now at 1.430v with prime95, i got 80C on hottest core just after 17 minutes, which is not good in my opinion, i will list my BIOS settings here for you to see if it can get any better:

PS. Windows Boots fine even on voltages as low as 1.28 with that frequency, but it all comes to Prime95

AI Tweaker:
AI Overclock Tuner: X.M.P.
Turbo (By All Cores): 47
Mem Freq.: 2133 (using Corsair Dominator GTs Quad C 2133 kit)
CPU VCore: Manual - 1.430v
CPU VTT: 1.15000v
VCCSA: Manual - 1.150v
DRAM v: 1.50 (xmp)
CPU & PCIE Spread Spectrum: Disabled

DIGI+ :
CPU LLC: High
CPU Current Capability: 120% (any higher and CPU power phase would lock to Extreme)
CPU Power Phase: Optimized
VCCSA LLC: High
VCCSA Current: 120%
VTT Over-current protection: Enabled
DRAM Current: 120% (all channels)
DRAM Power Phase: Optimized (all channels)

CPU Config:
Intel Adaptive Thermal: Disabled
DCU Streamer Prefetcher: Disabled
DCU IP Prefetcher: Disabled
CPU Power Management: All Enabled
Anti Surge Support: Disabled

All unlisted settings are left to default or auto.

Hope u can help me achieve a less aggressive voltage with some expertise help..

Here's a picture of my loop (don't mind the small cables, they are for the 2 180mm sidepanel fans).

[Arne_Saknussemm]

Hi there 🙂

Actually at first glance that looks about right to me...settings wise. Temp wise that is out of my comfort zone.

ANY BIOS version changes....Have you been running manual voltage all that time or offset? reason I ask is if you have been running a constant 1.4+ volts through the cpu with no/not much rest this may be the chip degrading....:(

[John93]

Arne Saknussemm wrote:
Hi there 🙂

Actually at first glance that looks about right to me...settings wise. Temp wise that is out of my comfort zone.

ANY BIOS version changes....Have you been running manual voltage all that time or offset? reason I ask is if you have been running a constant 1.4+ volts through the cpu with no/not much rest this may be the chip degrading....:(

I have been running manual voltage the whole time, & i was afraid that it's degrading, but then again, why would it degrade if the temps are always below 75 on all loads (except for prime95), cause only high temps could cause the CPU to degrade i guess. I've had decent temps even with my previous Corsair H80..

[Myk_SilentShado]

John93 wrote:
I have been running manual voltage the whole time, & i was afraid that it's degrading, but then again, why would it degrade if the temps are always below 75 on all loads (except for prime95), cause only high temps could cause the CPU to degrade i guess. I've had decent temps even with my previous Corsair H80..

CPU's degrade no matter how hard you push or don't push them. Even without OCing your CPU is still degrading, just at a much slower rate than an OC'd CPU will.

[Arne_Saknussemm]

Myk SilentShadow wrote:
CPU's degrade no matter how hard you push or don't push them.

This is quite true. The higher the voltage you put through it and the longer you do this, the quicker it will degrade.

In manual mode the voltage is constantly at the top level whereas using offset will allow your CPU to idle at lower voltage. Higher voltage will slowly kill your CPU so keeping it there whenever it is on is not a great thing to do I would guess. However 5 months may be a bit premature to assign blame to degradation (maybe not though) maybe BIOS change is to blame but you should be able to revert to your previous BIOS to see if it is down to that.

[John93]

I'm afraid it could be the new 2003 BIOS update cause as i remember, it didn't happen before that 😕

[John93]

I downed the clocks to 4.6GHz & vcore to 1.365, working fine now, 69c on hottest core, I just hope my CPU hasn't degraded massively somehow..

[Zka17]

You're having to high temps for a watercooling system! There is something wrong with it... Not sure if the cooling is causing your BSODs, but definitely you have to fix that issue too! Reseat your waterblock/repaste it, check the coolant's flow rate, check the airflow through the radiator...

What are the BSOD codes you're getting?

[John93]

Zka17 wrote:
You're having to high temps for a watercooling system! There is something wrong with it... Not sure if the cooling is causing your BSODs, but definitely you have to fix that issue too! Reseat your waterblock/repaste it, check the coolant's flow rate, check the airflow through the radiator...

What are the BSOD codes you're getting?

How can i check flow rate ? & i don't think this is too strange because temps barely reached 50C on 100% load at stock, my 7970 at 1125mhz core and 1575mhz memory gets to about 44C max, but i've bumped up the GPU voltage to the max and got max stable OC and now it tops at 61C with battlefield 3, & about BSOD codes, if i get it once again, i'll take a picture and take a look at the codes.

[HiVizMan]

What can happen is that dust builds up between the fins of the rad. Get a can of compressed air and give the rad and fans a good clean. But yeah 80'c are a tad hight.