First watercooling build - strange results

[BampirU]

Hello everybody,

This will be a very long post but I am facing some very strange results from my watercooling loop. This is my first watercooling loop so I am just going to show you the data I gathered and please tell me if you see anything wrong.

 

My main issue is that I am seeing very underwhelming cooling performance for my CPU while my GPU has quite good results. 

 

My Setup:
CPU: Intel i5 8600K
Mobo: MSI Z370 Krait Gaming
RAM: 16 GB DDR4 3000 MHZ G.Skill Dual Channel
GPU: Asus RX Vega 56 Strix Gaming OC
Sound card: Asus Xonar Strix Soar
SSD: Samsung 960 PRO, Intel 545s, Intel 660p (all M.2)
Case: Cooler Master Mastercase Pro 5
Fans: 6X Noctua NF-F12 chromax
PSU: Corsair RM650X

 

Watercooling components:
-        CPU Block: EK Supremacy MX, used with Noctua thermal paste (new tube)
-        GPU Block: EK-FC Radeon Vega Strix – Nickel, used with Noctua thermal paste and Thermal Grizzly pads
-        Rads: EK-Coolstream SE 240 & EK-Coolstream Classic PE 360
-        Pump combo: EK-XRES 100 Revo D5 PWM
-        Tubes: EK-Duraclear 9.5/12.7
-        Coolant: EK-Cryofuel Clear
-        Fittings: various EK-ACF 10/13mm

 

Please see below the temps I recorded (all after 15 mins full stress test on CPU & GPU, in 25C ambient):

Stock CPU + Stock GPU:
- Fans @ 40% & Pump @ 40%: CPU - 75; GPU - 50
- Fans @ 100% & Pump @ 40%: CPU - 66; GPU - 41
- Fans @ 100% & Pump @ 100%: CPU - 66; GPU - 38

 

5Ghz OC CPU + Stock GPU:
- Fans @ 40% & Pump @ 40%: CPU - 80; GPU - 52
- Fans @ 100% & Pump @ 40%: CPU - 75; GPU - 38
- Fans @ 100% & Pump @ 100%: CPU - 75; GPU - 38

 

For reference, while I was using air cooling, the 5 Ghz OC CPU temp was 80C using Noctua NH-D15 with 3X 140mm fans. Also, for my Strix GPU, max air cooled temp was 81C.
Idle temps are 33C for the CPU and 28C for the GPU.

 

2 thing I find really strange: 
- The difference between the OC temps on the Noctua NH-D14 and the watercooling loop is very small, was really expecting a big difference. Is is even stranger since the GPU lost 35-40 degrees going from air to watercooling. 
- In the same loop, I observe a 25C difference between CPU & GPU, this dose not seem normal to me

 

I’ve done the classic troubleshooting steps:
-        Removing the CPU block and making sure I have good contact and good spreading of the thermal paste
-        Only using a pee size ball of thermal paste in the middle of the CPU
-        Removing all air pockets from the loop
-        I even changed coolant, thinking I had a bad batch

 

I don't have a temperature sensor for the cooling liquid but I can confirm that the tubes are hot to the touch under full stress tests, no idea if it is normal and the 3 exhaust fans are absolutely vomiting hot air thought the top and back!
 
Please tell me if I am doing something wrong or if I am interpreting the data incorrectly. I know the GPU is in a X8 slot but from my testing and the results I found online, I can find no difference in performance for my GPU (same FPS in games, same results in benchmarks).

 

Thanks!

 

P.S.: sorry for the shit photos, literally took them at night with the phone under the desk.

[Derrk]

what voltage is your cpu running at?

[BampirU]

6 minutes ago, Derrk said:

what voltage is your cpu running at?

1.264v

[Derrk]

2 hours ago, BampirU said:

1.264v

id recommend delidding the cpu if you want better temps. you could also try lapping the top of IHS, but temp changes might be insignificant. 

cpu temps are safe as is, but seem a little warm for a custom loop with 2 rads. gpu temps seem where they should be

 

if you're worried about liquid temp, you should fill your reservoir to the top. it'll at least make the fluid take a little longer to reach max temps

[bignaz]

How did you have your NH-D15 cooling your gpu to?  Because you can't really compare a heatsink with a 103w load to a loop with an extra what 200w  from the Vega?

 

 

If you want super low cpu temps you gotta do what I do. Cpu gets it's own loop.  Even Idle your putting about 25 more watts of heat into that loop and on your cpu. 

 

How's your flow? And what's your temps when the loop is fully heatsoked. After an hour when your loop temps plateaus? Any bubbles? Shakenthe shit out of your raids after you filed your loop?  Also you could have a shitty ihs. Delid/lap can help. 

[BampirU]

13 minutes ago, bignaz said:

How did you have your NH-D15 cooling your gpu to?  Because you can't really compare a heatsink with a 103w load to a loop with an extra what 200w  from the Vega?

 

 

If you want super low cpu temps you gotta do what I do. Cpu gets it's own loop.  Even Idle your putting about 25 more watts of heat into that loop and on your cpu. 

 

How's your flow? And what's your temps when the loop is fully heatsoked. After an hour when your loop temps plateaus? Any bubbles? Shakenthe shit out of your raids after you filed your loop?  Also you could have a shitty ihs. Delid/lap can help. 

Obviously the D15 was on the CPU.

 

Unfortunetly I don't know how to check my flow, I don't have a flow meter in my loop yet. I don't see any bubbles, no matter how long or hard I use the computer.

 

Was afraid people would say deliding the CPU...