5950X Temperature Question

[DarkeVortex]

As far as what settings I'm running the CPU at - I'm not familiar with Ryzen OC. All I did was go into the Extreme Tweaker settings tab (ASUS BIOS), and change the AI OC from auto to DOCP Standard. That seems to get my RAM up to the right speed, etc, but does it change the CPU performance at all?

 

Once temps are more normal I'll dive into actual OC stuff...

[freeagent]

57 minutes ago, YoungBlade said:

What number are you using?

Good morning.

 

I woke up, powered up the machine and let it sit for a bit while I did my thing. Came back and fired up some OCCT and let it run for a half hour, and this is the result. It is 19c in the basement right now, and I am using 3 case fans, and 2 CPU fans.. she is not very quiet when she is busy lol..

 

[YoungBlade]

1 hour ago, freeagent said:

Good morning.

 

I woke up, powered up the machine and let it sit for a bit while I did my thing. Came back and fired up some OCCT and let it run for a half hour, and this is the result. It is 19c in the basement right now, and I am using 3 case fans, and 2 CPU fans.. she is not very quiet when she is busy lol..

 

I don't think that's a very good indicator for load temps, at least not compared to things like Cinebench R23. Running with those settings (AVX2, Large, Extreme, Variable, 24 threads) I reached steady state on Tdie at around 60C - likely because it's using AVX2. Regular AVX was the same, 60C. Setting it to use SSE raised steady state temp to 64C, probably because the CPU could boost a bit higher, but even that's not as high as Cinebench. Cinebench reached steady state at Tdie around 67C on my machine.

 

However, this is an apples and oranges comparison anyway, because I'm not overclocking. I'm actually just using Curve Optimizer to undervolt with an offset of -28, so my temps should be lower anyway.

[DarkeVortex]

This is a pic of my thermal paste. Looks super thin in the middle but squeezed around the outside- what did I do wrong?

[CommanderAlex]

3 minutes ago, DarkeVortex said:

This is a pic of my thermal paste. Looks super thin in the middle but squeezed around the outside- what did I do wrong?

~Snipped Photo~

Perhaps the coldplate or IHS is not completely flat and squeezed the thermal paste out on the sides...or you just overtightened the block on. 

 

On a side note, does your debug LED display your CPU temperatures once booted? Mine does but it always seems to be 10 degrees lower from what software is reporting. Just wanted to ask since I have a 5950X and X570 Dark Hero too. 

[DarkeVortex]

1 minute ago, SpiderMan said:

Perhaps the coldplate or IHS is not completely flat and squeezed the thermal paste out on the sides...or you just overtightened the block on. 

 

On a side note, does your debug LED display your CPU temperatures once booted? Mine does but it always seems to be 10 degrees lower from what software is reporting. Just wanted to ask since I have a 5950X and X570 Dark Hero too. 

Tbh, I haven’t checked the debug led. Once I got it posted I saw the temps and freaked out- didn’t even realize the debug LED showed anything other than post codes so thanks for letting me know.

 

I was hoping you wouldn’t say that about the block and IHS- not really in the mood to try and lap them… maybe overtightening was the problem? First time putting a monoblock on.

 

Gonna go do other stuff and wait for a few more replies and ideas before reassembling everything. Have to wait for the new valve anyway…

[CommanderAlex]

3 minutes ago, DarkeVortex said:

Tbh, I haven’t checked the debug led. Once I got it posted I saw the temps and freaked out- didn’t even realize the debug LED showed anything other than post codes so thanks for letting me know.

 

I was hoping you wouldn’t say that about the block and IHS- not really in the mood to try and lap them… maybe overtightening was the problem? First time putting a monoblock on.

 

Gonna go do other stuff and wait for a few more replies and ideas before reassembling everything. Have to wait for the new valve anyway…

Yeah no problem on having the temperature show after POST codes!!! I honestly forgot where it was in the BIOS as this is my first ASUS motherboard but I believe it was in a drop down on the top. 

 

I'm not sure if the monoblock came with instructions on how to tighten but perhaps they list a torque spec for the screws so you don't overtighten. 

 

For me, my 5950X and NH-D15 is around 40^oC at idle (3600MHz all-core boosting using PBO) with background processes and a game open in the background right now. Under load, it's below 80^oC in Cinebench R23 runs. 

[DarkeVortex]

4 minutes ago, SpiderMan said:

Yeah no problem on having the temperature show after POST codes!!! I honestly forgot where it was in the BIOS as this is my first ASUS motherboard but I believe it was in a drop down on the top. 

 

I'm not sure if the monoblock came with instructions on how to tighten but perhaps they list a torque spec for the screws so you don't overtighten. 

 

For me, my 5950X and NH-D15 is around 40^oC at idle (3600MHz all-core boosting using PBO) with background processes and a game open in the background right now. Under load, it's below 80^oC in Cinebench R23 runs. 

No torque spec (EK Crosshair VIII hero block, compatible with the dark hero). Just says “don’t overtighten”… thanks EK.

 

Im assuming what happened is that I put it on and tightened it, causing the thermal paste to spread, and then overtightened it, causing the 4 corners to get pulled down and the middle of the cold plate to bow up? Is that possible? Causing the air gap and awful temperatures I was getting.

[CommanderAlex]

Just now, DarkeVortex said:

No torque spec (EK Crosshair VIII hero block, compatible with the dark hero). Just says “don’t overtighten”… thanks EK.

 

Im assuming what happened is that I put it on and tightened it, causing the thermal paste to spread, and then overtightened it, causing the 4 corners to get pulled down and the middle of the cold plate to bow up? Is that possible? Causing the air gap and awful temperatures I was getting.

Or just perhaps improper tightening of the coldplate.

 

I can imagine in my mind what you are describing for the cold plate to bow up possibly. Good way to check if the surface is uneven is if you have a metal flat surface such as one found on a table saw with a cast iron top, lay the coldplate so that is in direct contact with the table and shine a light at the sides and see if any light comes through. If so, then you know the coldplate is not flat. 

 

I would just really redo the whole application of the thermal paste and tighten it back up more carefully and see if the temperatures improve. 

[DarkeVortex]

2 minutes ago, SpiderMan said:

Or just perhaps improper tightening of the coldplate.

 

I can imagine in my mind what you are describing for the cold plate to bow up possibly. Good way to check if the surface is uneven is if you have a metal flat surface such as one found on a table saw with a cast iron top, lay the coldplate so that is in direct contact with the table and shine a light at the sides and see if any light comes through. If so, then you know the coldplate is not flat. 

 

I would just really redo the whole application of the thermal paste and tighten it back up more carefully and see if the temperatures improve. 

Yep. That's the plan... after lunch  I think another mistake I made is that I installed the screws that hold the mosfet section of the monoblock first (this is the order the instructions give) but those are along 2 sides (top and left). Then I installed the 4 screws around the CPU after that (using the star pattern). But, next time I'll probably install the 4 CPU screws first to make sure the cold plate is mounted nice and flat against the CPU, then i can put on the screws for the mosfet portion afterwards since those use thermal pads anyway...

[DarkeVortex]

Spent most of the day rebuilding, reseated the monoblock, etc.

 

Just booted it up - same results. I'm at a loss...

[DarkeVortex]

On 8/14/2021 at 7:02 AM, YoungBlade said:

What number are you using? In HWInfo64, CPU (Tctl/Tdie) is consistently 20C over ambient at idle, but the individual CCD (Tdie)s are just 10C over ambient. CPU Die (average) is 15C over ambient.

 

EDIT: I decided to check "true" idle temps for the CPU - as in nothing running except Windows and HWInfo64, closing all other applications. Before, I was looking at temps with Firefox running, which may have kept some of the cores active - who knows what the browser is doing in the background.

 

I let the computer sit until temps seemed to stop dropping, then pressed the reset button on HWInfo64, and waited 3 minutes. Then, I proceeded to record the numbers on some pressed plant fibers via a dark liquid on a stylus so as not to disturb the computer (like how the Ancient Egyptians recorded their data - it's crude, but effective). Here's what I got:

 

CPU (Tctl/Tdie) - Min: 33.4C Max: 34.8C Avg: 34.0C

CPU Die (Average) - Min: 29.1C Max: 31.3C Avg: 29.6C

CCD1 (Tdie) - Min: 26.0C Max: 42.0C Avg: 27.5C

CCD2 (Tdie) - Min: 28.5C Max: 38.3C Avg: 30.8C

Core Temperatures - Min: 24.0C Max: 30.2C Avg: 26.2C

 

It got rather cold here last night - 57F (14C) - and I opened up the windows. Obviously, it's not that cold in here, but it did get down to 65F (18C). So Tdie is 16C over ambient. CCD (TDie)s are about 11C over ambient. And Die (Average) is 12C over ambient. So the CCD (TDie)s stayed the same, while Tdie and Die (Average) dropped a few degrees by closing Firefox.

 

With Intel, I regularly saw average temps in the low 20C range when it got that cold in here. But I don't think this means that Ryzen runs hot. I think it means that AMD reports differently from Intel.

Just did this test. Here's what I got:

TAmbient: 27.0C

CPU (Tctl/Tdie) - Min: 41.1C Max: 71.1C Avg: 47.5C

CPU Die (Average) - Min: 37.8C Max: 71.2C Avg: 42.9C

CCD1 (Tdie) - Min: 40.5C Max: 74.0C Avg: 46.6C

CCD2 (Tdie) - Min: 36.8C Max: 67.0C Avg: 40.7C

Core Temperatures - Min: 35.1C Max: 72.4C Avg: 39.1C

GPU Temp - Min: 30.9C Max: 34.2C Avg: 31.3C

Fluid Temp - 35C

 

looks like some random process started up that caused a quick spike on the CPU, but didn't last long. Didn't see what it was, unfortunately. I'm scared to try a test under load if these are my "idle" temps, since I don't want to do any damage...

 

EDIT: This was also with my pump on max RPM, and all 9 fans (3 360rads) at max RPM (~1600)

[freeagent]

Seems warm for water cooling, especially for 3x rads.. but I know nothing.

[DarkeVortex]

2 minutes ago, freeagent said:

Seems warm for water cooling, especially for 3x rads.. but I know nothing.

Yeah. Just trying to figure out what’s going on, why there would be bad contact with the water block even after a reseat… unless I’m just totally inept

[DarkeVortex]

Here's a snip after a 10min multicore Cinebench R23 run. T_Ambient: 27C, Cinebench Score: 28209pts (Multi core)

[YoungBlade]

1 minute ago, DarkeVortex said:

Here's a snip after a 10min multicore Cinebench R23 run. T_Ambient: 27C, Cinebench Score: 28209pts (Multi core)

That's possibly warm considering that you have 3 360mm rads, but 27C ambient is also on the high side. Your score isn't far off the ones I find online on sites like cpu-monkey, so you aren't losing much performance. You also aren't breaking 90C on Tdie, so you aren't getting into throttling territory, but it's close.

[DarkeVortex]

5 minutes ago, YoungBlade said:

That's possibly warm considering that you have 3 360mm rads, but 27C ambient is also on the high side. Your score isn't far off the ones I find online on sites like cpu-monkey, so you aren't losing much performance. You also aren't breaking 90C on Tdie, so you aren't getting into throttling territory, but it's close.

Yeah. My worry is that with the CPU running and the GPU actually kicking in, it'll pass 90 as the water warms up more. Maybe it's just from the room being hot (I ofc have to be set up in the hottest room of the house where the AC barely reaches) but it just still feels wrong, especially under water...

[Guest]

Sharing VRM heat with CPU heat on the same cold plate.

 

Plus warm ambient (this part is important)

At 80f-27c

 

I would be curious to experiment using a fan at the left and top VRM sections of the board. 

Wondering if a little air flow under the mono block would help.

 

Have seen at various forums in the past actually cooling the back side of the board at the cpu. We see thermal transfer in the tracings as well.

[DarkeVortex]

4 minutes ago, ShrimpBrime said:

Sharing VRM heat with CPU heat on the same cold plate.

 

Plus warm ambient (this part is important)

At 80f-27c

 

I would be curious to experiment using a fan at the left and top VRM sections of the board. 

Wondering if a little air flow under the mono block would help.

 

Have seen at various forums in the past actually cooling the back side of the board at the cpu. We see thermal transfer in the tracings as well.

Interesting. Might be something worth trying. Do you not think these temperatures are exceptionally high, even with the high ambient temp?

[freeagent]

The high ambient explains it.

[Guest]

Just now, DarkeVortex said:

Interesting. Might be something worth trying. Do you not think these temperatures are exceptionally high, even with the high ambient temp?

I do not think the temps are exceptionally high. It is mostly a result of 27c ambient temp. 

Where I'm at, the room temp is 66f atm. And at times can be over 80f like yours. Just depends on the weather and how hard I run the AC in the house. But we've been having nice cool evenings, so AC stays off till early after noon.

 

About the cooling and the fans theory.

 

You are only cooling a single surface. Most of the heat is dissipated here. 

 

If the block has an acrylic top, this is an insulator. You cannot use a fan to dissipate heat and remove additional BTU from the block it's self. So again, only transfer thermals in a single direction. 

 

And why I recommend full copper water blocks and actively cool VRMs with fans which will also aid in removing thermals from the water block. 

 

Also you have a beastly processor. So not exactly the easiest thing to cool in the first place.

[DarkeVortex]

1 minute ago, freeagent said:

The high ambient explains it.

Really? I'd been running my 9700K under a 360 AIO forever in this warm temperature, just until this rebuild, and I felt like I had more headroom in general. I know this is a much denser / powerful chip, but wow

[DarkeVortex]

Just now, ShrimpBrime said:

I do not think the temps are exceptionally high. It is mostly a result of 27c ambient temp. 

Where I'm at, the room temp is 66f atm. And at times can be over 80f like yours. Just depends on the weather and how hard I run the AC in the house. But we've been having nice cool evenings, so AC stays off till early after noon.

 

About the cooling and the fans theory.

 

You are only cooling a single surface. Most of the heat is dissipated here. 

 

If the block has an acrylic top, this is an insulator. You cannot use a fan to dissipate heat and remove additional BTU from the block it's self. So again, only transfer thermals in a single direction. 

 

And why I recommend full copper water blocks and actively cool VRMs with fans which will also aid in removing thermals from the water block. 

 

Also you have a beastly processor. So not exactly the easiest thing to cool in the first place.

All good points. Definitely worried to run a CPU + GPU test to see what happens when the water warms up, but maybe I should hold off on that until it cools down a bit...

[Guest]

4 minutes ago, DarkeVortex said:

All good points. Definitely worried to run a CPU + GPU test to see what happens when the water warms up, but maybe I should hold off on that until it cools down a bit...

Fans. Put some fans on it. 

Quick, easy and cost effective. 

[Kinda Bottlenecked]

Do you have any of the asus motherboard software installed?