5930k on a X99 MSI Carbon water cooled 85c

tez

tez

tez

Associate
Joined
27 Nov 2005
Posts
18
Hi.

Just got my MSI X99 Carbon pro gaming board, 5930k cpu which is water cooled.

Looking for some advice regarding essencial things I maybe need to turn off in bios, and optimal settings for this combo.

Currently im at 4.5ghz 1.27v, I havent a clue what half of these bios settings are, but im booting into windows, and running AIDA64 System stability and its not showing errors, however Idle the cpu is 60c, load at 85c

My memory is running at its full speed 3200mhz, yet it didnt bring the base clock up to 125, it stayed at 100?

Thanks, Dan
 
That's unusually high for a proper watercooling loop... perhaps there are issues.

If it's a single-fan rad AIO cooler though - that sounds about right.

2x120 or 2x140 should do better than that though at those speeds and voltages.
 
My water cooling system is as follows.

AquaComputer Kryos Cluplex XT block 1/2" pipes
2x single 120 Rad XSPC Slim, 2x 120 fans sucking air out placed outside the case

2x 120 fans front of case pulling air into case
1x 120 fan next to cpu sucking air out the case, fan placed outside the case.

12v D5 pump.

Case TT Mozart TX case

Dan
 
Im using Artic Silver, very thin layer, I will check the spread just in case. I will also have a check for air in the system, thanks for the advice.

Dan
 
tez said:
Im using Artic Silver, very thin layer, I will check the spread just in case. I will also have a check for air in the system, thanks for the advice.

Dan
Click to expand...

The d5 is more than capable for your setup so cycle it between low and high speed whilst tilting the case in all angles... Eventually you should stop hearing it gurgle.
Keep us updated :)

andicole0 said:
Wow! biased or what!?
Andi.
Click to expand...

Why would I not be? The AIOs are inferior in every way... If price and/or convenience are your thing then get a decent air cooler. Just my experience having dealt with all 3 :)
 
I have no experience with the x99 stuff but maybe they do run that hot? Worth checking the TIM out though.

Sounds like you haven't got any airlocks so it should be pulling heat away from the chip nicely... Are the rads warm to touch?
 
NoNameNoNumber said:
I have no experience with the x99 stuff but maybe they do run that hot? Worth checking the TIM out though.

Sounds like you haven't got any airlocks so it should be pulling heat away from the chip nicely... Are the rads warm to touch?
Click to expand...

They don't run that... a system with 2x 120mm rads should be able to cool it better than that. Heck, a Noctua D14/D15 would do better than that... sounds like there is any issue to me.



To OP... just to check... I assume that loop is just for the CPU only and not the GPU too? If it's for both, that might explain temps.

If CPU-only... I assume you've checked fan profiles aren't turning the pump off or anything similar? You need to keep the pump running all the time and especially auto-fan-profile software can affect this.

Also check fans are running etc and try turning everything to 100% just for testing.

If that's all done and you're still seeing those temps... I would suggest that there is either a problem with the contact between the CPU and CPU watercooling block... so try re-mounting. Like someone else suggested, also check how well the TIM has coated both the CPU and block when you take it off.
 
Ok. I ran into some big problems last night. Wanting to get the best stability, I looked about on the net and found I could adjust the ring ratio and frequency at the suggested 35 to get it to 3500mhz from auto 3000mhz. I also set some manual voltage settings such as 1.350 for memory. I got into windows, was running Aida64, and was getting freezing. My rendering app crashes not far into a render. I started to use the MSI clocking tools in windows taking the cpu voltage up, and down testing for stability. After a restart I went back to a clock profile that iv been using all day.

When I ran AIDA64 it failed within 30 seconds. I put the bios to 4ghz, then back to optimized defaults but Aida64 failed every time. I then ran each test by itself, and it was failing of the Cache and memory stress tests. Then I ran memtest86, and got 14 errors within a 3 hour cycle. My system was still at 4ghz, but memory at stock settings

I tried to run test number 7 which had got me 9 errors before, but it didn't seem to show errors? I then again reset bios to optimized default, ran each of the Aida64 stess tests again, and they all passed? I ran my render app and cinebench, and no errors or crashes. I then ran Windows memory diagnostic tool, and it passed.

Now I'm not sure if something is faulty or damaged. Iv not clocked my memory at all. At the moment running all 4 Adia64 tests my cpu is not exceeding 55c, so If I was on air god knows what it would be being my cpu is on boost to 3.6ghz? Still so far looks like iv not broken anything, cant account for why last night it was so unstable although I clocked everything yo stock.

Dan
 
Last edited:
I am only using a corsair h110gtx AIO (Noctua Fans) with my 5960x at 1.32v (4.5ghz) and my temps are max 81 at load over 8 hours stable Aida64...So if you are using a full water loop surely you should have better results? I still have room to overclock but like already mentioned your temps don't seem correct.
 
Last edited:
My rooms temps are about 22c. I have 5x 120mm case fans running, three of which are Noctua Fans. I think its just the case that this chip runs hot, and if I didnt have water cooling there would be no chance of any overclock hardly. I have my cpu at 1.275v, 4.5ghz, and its hits 80c on load. Im rendering so the load is heavy and 100%.

Dan
 
Last edited:
Your getting 60 degrees idle? Something wrong there with 22 degrees room temp.

Are you rads warm really warm? If they are then turn your fans up.
If your rads are cold feel your cpu block, if thats hot then check your pump is running.
If your block is cold then clean the thermal paste off and reapply.

High idle temps could mean your just running stuff in the background, check CPU usage when idle, i think its more likely since you only have a 20 degree increase to load temps.
 
Something wrong there... even if you had the cpu running at 4.5GHz permanently (ie, not using turbo boost) then it still shouldn't sit that high at those volts with that cooling when idle.

Do what BluD says... if that doesn't help, check the CPU block mounting and thermal paste as previously suggested.
 
That's my thought... at 60C idle, it could easily be a fan profile that's turning off the pump when the temperature is under 60C - which is the normal-ish target spin-up temp for a lot of low-noise fan controlling software
 
You must log in or register to reply here.
Back
Top Bottom