3970x | rive - very high core temp difference - very high core temp while gaming!

[kingsaadi]

Hey guys,

I have been having issues with my 3970x cooled by Kraken X61 for a couple of months now.

I'm getting unusually high CPU load temps with cores 1 & 2 going above 80*C. This is while just gaming. On Battlefield 1 and Far Cry 4 so far I tried.

I initially thought it was the stock thermal paste as it had been there for almost 2 yrs so I reapplied fresh thermal paste - Noctua NTH1. As far as I know Noctual NTH1 doesn't have curing time but I still let it run for a couple of weeks before I took temps just to be sure.

I've checked CAM software that Kraken X61 uses to run the pump - made sure it's set to run by changes of CPU temp, not LIQUID temp. Fans are set to go at 100% after 60*. Made sure motherboard is giving 100% power to pump by setting CPU Qfan to 100% for both min and max temps.

Played Far Cry 4 for couple hours. Only uses first 2 cores mainly, with a little bit of core 3 but temps still hover between 75-80*, maxing out at 82*. Core 3 maxes out at 74*. Cores 1 & 2 hit 80+. These are way to high for just gaming temps.

I'm not sure why this is happening now as the problem wasn't there when I first purchased the Kraken. Temps were well below 75*C even while stress testing.

Also did a quick stress test running CPU burn in 1.01 - running 12 instances for my 3970x (6 cores 12 thread). Attached is result from Real Temp GT 3.7.

As you can see core 1 is 80*C. Cores 2, 3 - OVER 80 as well! Temperature difference between lowest and highest core is almost 20*C!

I've reapplied thermal paste and reseated the waterblock 3 times now - not much different.


Any help to figure out what's causing this problem will be greatly appreciated 🙂

PC SPECES:
3970x @ 4ghz
RIVE
Kraken X61 with 2xNoctua industrial A14 2000rpm fans
4x8GB 1866mhz DDR3 Corsair Dominators
2xGTX980 SLI
Intel 520series SSD
WD black 2TB HDD
Xonar essence STX sound card
NZXT phantom 820


Many thanks,

Saadi

[Arne_Saknussemm]

I would suspect the cooler has some problem...

Maybe uninstall it and see if you can hear liquid sloshing around in it...which would mean air bubbles...which there should not be...that sort of thing...

[kingsaadi]

Arne Saknussemm wrote:
I would suspect the cooler has some problem...

Maybe uninstall it and see if you can hear liquid sloshing around in it...which would mean air bubbles...which there should not be...that sort of thing...

Thanks guys,

I can't recall any liquid sloshing around last time I removed the unit. HOWEVER, there is a random 'sloshing' around of liquid while the cooler is installed and running. It's sometimes there when changing the 'mode' from 'silent' to 'performance' or vice versa. Sometimes I'd hear it randomly while doing nothing. It only lasts 1 second or so.

Is it possible that I tightened the thumbscrews too much when mounting the waterblock and could that cause problems??? I did diagonal cross pattern to make sure it was even. Tightened using fingers as much as I could then finished off with screw driver, only using my fingers to tighten as much as possible. There is no movement on the waterblock after installation so it does seem to have a good fit.

Are there any other indicators that the pump is broken??...like idle temps vs liquid temp difference or something (I'm just throwing out ideas here lol)?


Cheers

Saadi

[Menthol]

If you have used that cooler for 2 years I would say it is probably the pump giving out (even if the software shows the pump speed is normal), or as Arne says maybe evaporation and the system is low on liquid, not being able to service or replace parts is the downside to AIO coolers and a replacement may be needed, this happened to me and a Corsair h110i, temps started rising, everything seemed to function normally but a new cooler was the only cure.
2 years is probably a reasonable to long life span for any AIO

[Arne_Saknussemm]

You would have to check out the pattern in the TIM (thermal paste) to make sure contact is good between the block and the CPU....should be virtually see through even spread over CPU IHS surface when you remove the block. I would say never to use anything other than finger tight...finishing off with screw driver or pliers can displace CPU in socket or break socket.

So...you could try remounting...again...but the noise from the cooler pretty much makes me think it needs replacing. There should be no liquid sounds...no air bubbles...and idle temps should be very close to 30 degrees assuming a normal ambient temp.

[kingsaadi]

Thanks a lot people 🙂

Definitely was cooler issue. When I took out the unit I could hear water sloshing around as I took the unit out of the case. It went away after sloshing around a couple of times. Installed a air cooler and temps are back to normal.

I had to reset BIOS to optimised defaults and redo my xmp profile settings as the night before it randomly started to bsod while watching a youtube video. Got quite worried as it kept bsod-ing with x128 stop code. When analysed it kept saying 'probably caused by: genuine intel'.

---------------------------------
Another issue -

There seems to be an issue with cold boot posting. When I boot after long period of time the system doesn't post - I have to restart, then it posts. What could be the issue here?...Should I start another thread for that?

Thanks

Saadi

[Arne_Saknussemm]

OK...ship that cooler off for RMA...

the 0x128...Maybe a driver issue,,,?

Does the system not boot if it's at F5 defaults?

[kingsaadi]

Arne Saknussemm wrote:
OK...ship that cooler off for RMA...

the 0x128...Maybe a driver issue,,,?

Does the system not boot if it's at F5 defaults?

The bsods are gone after I set to defaults and uninstalled the Kraken driver/software - CAM.

Resetting to optimised defaults works fine (i assume that's the same as f5 defaults?). And just to clarify, the system is "Booting", but not "Posting".

However, since I'm running 32gb 1866 rams I set the profile to XMP and adjusted the ram volts manually to 1.65.

Some of the other settings I changed -

CPU spread spectrum - disable
PCIE spread spectrum - disable

Power limit control - Enabled (I changed this from "Auto" to "Enabled" because I noticed power limit throtlling while benchmarking. CPU would turbo at 4ghz for a short while before cutting back to 3.7ghz. This happened as power went past the TDP of 150w for about 10 seconds or so). I set the new values at:

- long duration power limit - 200
- short duration power limit - 240

The issue seems to be only with the first boot of the day. If I shut down the PC and turn it back on after 5-10 mins it boots and posts fine.

Managed to take a photo of the motherboard post code..sorry it turned out quite blurry. But you can see -

Post code 62,
Red light next to power connector.

Is there any settings in BIOS that I should modify?

Thanks!

[Arne_Saknussemm]

that's b2....memory training....is that a single quad channel kit of RAM?

[kingsaadi]

umm...it's 2 kits of 2x8gb..as in 1 package had 2x8gb of ram. So I bought 2 packages.

So, I guess its not a 'single kit' of quad channel, rather it's 2 kits of dual channel? It's running at quad channel according to cpuz.