Improving Cooling / Fans Upgrade In Clevo

[dude-137]

I'm looking for the ways to improve cooling on my Clevo P751TM1-G having i9-9900K + RTX 2080.

 

But the thread is useful for P75xTMx(-G)/P75xDMx(-G)/P775TMx(-G)/P870DM(-G)/X170KM-G/etc too, as the fans are physically the same in all of them.

 

The CPU in my Clevo is undervolted and configured to have the all-core frequency of 4.2GHz (sometimes having to set it to 4.1GHz).

Higher than that makes it run too hot.

Given that the CPU's default all-core frequency is 4.7GHz and can be overclocked all the way up to 5GHz, there's certainly a huge theoretical performance improvement potential.

 

Would be grateful for any recommendations / tips! :-)

 

In particular, I'm curious about the potential fans upgrade.

 

The default fans both on the CPU-side (Left) & GPU-side (Right) in P751TM1-G are:

ADDA, 6-31-P7753-100, 180925, 00DM3, DC 12V, 0.5A, 17 leaves

 

I'm quite sure there should be various different models of fans that are swap-in-compatible with the Clevo's default fans.

Has anyone upgraded their fans to more powerful ones?

If yes, could you share the model names?

 

Here's how I would prioritize power vs. noise:

1. Ideal case - a model that is both more powerful/better-cooling and quieter :-) 

2. A more powerful and better-cooling model, without caring about its noise.

3. A model that is the same as the Clevo's default fans in performance, but quieter.

 

 

PS:

Fans is the most obvious upgrade idea that comes in mind, but are there any other possible cooling-hardware update options or other ideas?

[kiemsigh]

I use water cooling

[Maro97]

what is the CPU frequency on all cores ? and how much CPU TDP . with cenabench r23 test ? 

[Maro97]

Here is my results with new nvidia driver 150w rtx 3080 with two different cpu 8700k & 9900kf with air cooling & UV 

[SuperMG]

26 minutes ago, Maro97 said:

Here is my results with new nvidia driver 150w rtx 3080 with two different cpu 8700k & 9900kf with air cooling & UV 

I get 270-280FPS with an RTX 4080 150W and i7 4900MQ bruh and 190FPS with an RTX 5000 (RTX 2080 100W) and i7 3940XM...

 

How much do you get for FurMark 1.37?

[Maro97]

I don’t benchmark all graphics cards by furmark I’m using furmark program only to test the GPU temperature and GPU hotspot for one hour . 

But for evaluating frame rates, I use 3DMark (Time Spy & Fire Strike) and the Heaven Benchmark

[Maro97]

2 minutes ago, Maro97 said:

I don’t benchmark all graphics cards by furmark I’m using furmark 

program only to test the GPU temperature and GPU hotspot for one hour . 

But for evaluating frame rates, I use 3DMark (Time Spy & Fire Strike) and the Heaven Benchmark

As you can see in the image, the GPU temperature does not exceed 65°C, and the difference between the hotspot and the GPU temperature is about 6°C, which also confirms that the thermal pads are installed with the correct thickness

[Maro97]

This is exactly what I do every time I test a new thermal paste that has recently been released on the market.

For gpu cooling I’m using a thermal grizzly duronaut. 

[SuperMG]

33 minutes ago, Maro97 said:

This is exactly what I do every time I test a new thermal paste that has recently been released on the market.

For gpu cooling I’m using a thermal grizzly duronaut. 

Not this Furmark. The Furmark 1.37 you can download at geeks3d, there's a one minute benchmark at 1080p.

 

What about 3DMark TimeSpy?

[Maro97]

3 hours ago, SuperMG said:

Not this Furmark. The Furmark 1.37 you can download at geeks3d, there's a one minute benchmark at 1080p.

 

What about 3DMark TimeSpy?

 

[SuperMG]

1 hour ago, Maro97 said:

 

I get 16000 graphics score with the 4080 150W MXM and 17700 with the 4090 130W MXM. 12400 with the RTX 4070 MXM and i7 7th gen. All running at PCIe 3.0 (expect for the 4090, at 2.0 x16)

[Maro97]

5 hours ago, SuperMG said:

Not this Furmark. The Furmark 1.37 you can download at geeks3d, there's a one minute benchmark at 1080p.

 

What about 3DMark TimeSpy?

 

1 hour ago, SuperMG said:

I get 16000 graphics score with the 4080 150W MXM and 17700 with the 4090 130W MXM. 12400 with the RTX 4070 MXM and i7 7th gen. All running at PCIe 3.0 (expect for the 4090, at 2.0 x16)

4090 with ZRT VBIOS ?

[SuperMG]

27 minutes ago, Maro97 said:

 

4090 with ZRT VBIOS ?

No ZRT VBIOS on these tests. Because the ZRT ones are at 110W max. I don't know how they're able to advertise as a 150W GPU. The performance is worser. @DynamiteZerghas a ZRT 4090 card. I only tested the vbios in a X-Vsion PCB card. I'm waiting for his answer about his ZRT PCB card.

[Maro97]

5 hours ago, SuperMG said:

I get 16000 graphics score with the 4080 150W MXM and 17700 with the 4090 130W MXM. 12400 with the RTX 4070 MXM and i7 7th gen. All running at PCIe 3.0 (expect for the 4090, at 2.0 x16)

Which laptop tested with 4080 / 4090 ?

[dude-137]

On 4/27/2026 at 5:07 PM, dude-137 said:

Has anyone upgraded their fans to more powerful ones?

If yes, could you share the model names?

 

Here's how I would prioritize power vs. noise:

1. Ideal case - a model that is both more powerful/better-cooling and quieter :-) 

2. A more powerful and better-cooling model, without caring about its noise.

3. A model that is the same as the Clevo's default fans in performance, but quieter.

 

Has anyone done this?

[dude-137]

What do you think about this, are there any potential negative sides?

 

So far I see only the positive ones:

- Significant simplification of dust cleaning

- Significant reduction or almost complete removal of dust reaching and sticking to the fans (only some amount of the smallest particles make it through)

- No need for any physical interaction with the fans and their surroundings

- The fans keep being "young" for much longer

- Thus, the default cooling efficiency stays the same much longer and doesn't degrade over the time

[SuperMG]

On 5/1/2026 at 10:21 AM, Maro97 said:

I don’t benchmark all graphics cards by furmark I’m using furmark program only to test the GPU temperature and GPU hotspot for one hour . 

But for evaluating frame rates, I use 3DMark (Time Spy & Fire Strike) and the Heaven Benchmark

For the RTX 4080 with the custom made 3 fat heatpipes for the P775TM: I got 368FPS in 1080p Ultra with an i9 9900KF (-75mV undervolt and 5GHz on 4 cores).

 

I get 280FPS with an i7 4900MQ (4.1GHz on 2 cores)

 

Furmark both systems have the same score. TimeSpy too, same graphics score of 16k

[SuperMG]

Hello.

 

I got 19700 points in TimeSpy graphics and 386FPS in heaven benchmark with the RTX 4090 150W in the P775TM1.

 

With a P570WM3 and 4090 130W I got 17700 graphics points.

[SuperMG]

Hello everyone.

 

How can I get 11k in CPU score in TimeSpy?

 

I can get max 9.8k. the average clock speed says 4.59GHz and not higher. All the cores are set to 4.9GHz with a -75mV undervolt.

 

My heatsink is a stock 3 heatpipes P775 one. 2 fat and 1 normal.

 

If I add one extra heatpipe

or

 

If I replace the third small heatpipe with a wider one... will it improve the performance and I can reach 11k or not?

[Kastner]

4 hours ago, SuperMG said:

How can I get 11k in CPU score in TimeSpy?

 

I can get max 9.8k. the average clock speed says 4.59GHz and not higher. All the cores are set to 4.9GHz with a -75mV undervolt.

Hi,

 

What is your Pl1 and Pl2 level? With unlimited you should have around 11k.

[SuperMG]

Just now, Kastner said:

Hi,

 

What is your Pl1 and Pl2 level? With unlimited you should have around 11k.

180W on both and 1023A current