Minisforum MS-01 PCIe Card and RAM Compatibility Thread

Notice: Page may contain affiliate links for which we may earn a small commission through services like Amazon Affiliates or Skimlinks.

00riddler

New Member
May 14, 2017
11
1
3
41
In theory you could use the full 115W. Not sure if the cooler can handle that on the long term.
I think they also limited it so that there is some headroom for a potential graphics card.
In the end you are limited by the cooler and your power adapter.
 

wadup

Member
Feb 13, 2024
98
66
18
In theory you could use the full 115W. Not sure if the cooler can handle that on the long term.
I think they also limited it so that there is some headroom for a potential graphics card.
In the end you are limited by the cooler and your power adapter.
So in theory if you have a 240W power adapter and cooling is good you should be able to turn it all off?
 

caplam

Member
Dec 12, 2018
58
12
8
I'm running memtest as i have to wait for a hba to finish my set-up. It's running quite hot.
There is only the ram, no ssd ,no pcie card.
Do that seem normal to you ?
Capture d’écran 2024-03-01 à 11.31.25.png

edit: i forgot to mention the ram used for ther purpose of this topic:
crucial 2x32GB kit DDR5-4800 CT2K32G48C4085

I'm waiting for a lsi 9300-8e hba which i think it should work.
 

heromode

Active Member
May 25, 2020
380
203
43
Is it too soon to suggest an 3D printed alternate case, bigger fans and more?
I still think these high performance components crammed into a small case, and cooled with small, custom sized fans is not a viable solution. I'd like a "Tower Cooler" like design where one or two say 200mm casefans would be mounted outside the case, on one or both sides (push - pull), just like tower coolers. The case would be perforated with airfunnels that would direct the airflow to important components. Maybe the CPU would still need a internal fan.. Anyway with externally mounted standard casefans, it would be possible to change the fans even with the system running etc..
 

skipper ohms

Member
Jan 24, 2024
35
23
8
So are these settings low due to Minisforum being overly cautious etc?
Nice to know PL2 isn't locked down by Minisforum in these, thanks!

If you can run at max load without reboots or thermal issues, probably good to go. Although sometimes a manufacturer will use a junky VRM or ship a weaker PSU. Here's what a robot has to say about VRMs, ymmv:

>>> user

if i set PL2 too high on a weak VRM, will the VRM melt?

<<< assistant

Setting PL2 (Power Level 2) too high on a motherboard with a weak VRM (Voltage Regulator Module) can potentially lead to overheating and failure of the VRM if it is not designed to handle the sustained power delivery. While "melting" is an exaggeration, the VRM can overheat, which might cause the system to throttle performance, shut down to prevent damage, or in extreme cases, fail completely. It is important to ensure that the VRM is capable of handling the power demands of the CPU, especially under PL2 conditions, which typically occur during short-term turbo boost scenarios. Adequate cooling and proper case ventilation are also crucial to maintain VRM and overall system health.

>>> user

how do i see VRM temps in linux?

<<< assistant

To see VRM temperatures in Linux, you can use the `lm-sensors` package. Here's how to install it and check temperatures:

1. Install `lm-sensors`:
```bash
sudo apt update
sudo apt install lm-sensors
```

2. Detect your sensors:
```bash
sudo sensors-detect
```
Follow the prompts and agree to detect the sensors.

3. Check the temperatures:
```bash
sensors
```

The output will list temperatures for various components. VRM temperatures might be listed as something like "VRM Temp", "PCH Temp", "Motherboard Temp", or similar, depending on your hardware and sensors available. If VRM temperatures are not listed, your motherboard might not have the necessary sensors, or they might not be supported by `lm-sensors`.

>>> user

what's max VRM temp?

<<< assistant

The maximum VRM temperature varies depending on the manufacturer and model of the motherboard or VRM components. Typically, VRMs are designed to handle temperatures of around 90°C to 125°C. However, for sustained operation and longevity, it is generally recommended to keep VRM temperatures below 85°C. Always refer to the specifications provided by the manufacturer for the maximum safe operating temperature of your specific hardware.
 

wadup

Member
Feb 13, 2024
98
66
18
So for those of you removing the wifi card and not replacing it with anything. I just removed mine from the pcie slot and rested it on the motherboard rather than removing the tape and unplugging the antenna connectors. Those antenna connectors are so hard to plug back in sometimes.
 
Last edited:

wadup

Member
Feb 13, 2024
98
66
18
So I was playing around with fan speeds in the bios for the cpu. I picked up some nice temp drops without any audible difference on idle. My high loads are bursts and MS-01 is in a closet so I don't hear it. I don't mind the fan being loud when its working hard. At idle I felt like the fan was too slow so I wanted to see how high I could turn it up and still be silent.

Here are the stock settings:
Screenshot 2024-03-01 111817.png

This is what I settled on:
Screenshot 2024-03-01 103434.png

My idle temps dropped 4C.

There are some definite improvements to be made facturing in your climate, noise levels and fan speeds.
 

GreenAvacado

Active Member
Sep 25, 2022
170
66
28
So I was playing around with fan speeds in the bios for the cpu. I picked up some nice temp drops without any audible difference on idle. My high loads are bursts and MS-01 is in a closet so I don't hear it. I don't mind the fan being loud when its working hard. At idle I felt like the fan was too slow so I wanted to see how high I could turn it up and still be silent.

Here are the stock settings:
View attachment 35102

This is what I settled on:
View attachment 35103

My idle temps dropped 4C.

There are some definite improvements to be made facturing in your climate, noise levels and fan speeds.
That is neat. Only downside to this approach is fan running 24x7 and will age it much quickly. Personally I prefer getting away with as much passive cooling until it no longer works but then again, I'm sitting in the same room.
 
Last edited:

wadup

Member
Feb 13, 2024
98
66
18
That is neat. Only downside to this approach is fan running 24x7 and will age it much quickly. Personally I prefer getting away with as much passive cooling until it no longer works but then again, I'm sitting in the same room.
Yeah if I was in the same room I would take the same approach as you.
 

andrei4002

New Member
Feb 15, 2024
7
0
1
is there any recommendation for a pcie card that would fit in there and support another 2 NVMe drives (or more)?
 

Fazio

New Member
Dec 20, 2022
28
21
3
Just replaced the thermal paste on my MS-01. I'll monitor during the next days but first data seems great!
1709389759059.png
Core package is stable at 40°C with the same load/environment as before.

Green is CPU package, yellow is NVME (990 Pro), red is ambient temperature.
I used ARCTIC MX-6.
 

jmoschetti45

New Member
Feb 24, 2024
2
0
1
Radeon RX 6400 seems to be the go to confirmed working GPU choice for most people, but the 1x HDMI 1x DP isn't what I ideally need. I stumbled across the PRO W6400 which seems to be identical minus having 2x DP which is exactly what I need. Any bets on this working or not? Want to avoid an HDMI > DP converter if at all possible.
 

wadup

Member
Feb 13, 2024
98
66
18
At least -15°C from the stock paste so far. I had 55-60 with very low usage, now 40-45. I’ll monitor it during the weekend.
I would contribute the 15C more to a cold boot after paste application then anything else. For a true test you really want to let the machine warm up then put a consistent load on your CPU stock vs mx6/LM. I doubt your improvement will be 15C with new paste usually it's around 5C.

The thing you have to realize is Thermal Grizzly Conductonaut Extreme conductivity of 79W/mK while MX6 conductivity of 7.5W/mK. That is a big difference and I don't want people on here thinking they are going to get 15C from premium paste.
 

wadup

Member
Feb 13, 2024
98
66
18
Radeon RX 6400 seems to be the go to confirmed working GPU choice for most people, but the 1x HDMI 1x DP isn't what I ideally need. I stumbled across the PRO W6400 which seems to be identical minus having 2x DP which is exactly what I need. Any bets on this working or not? Want to avoid an HDMI > DP converter if at all possible.
Is this the card? https://www.amd.com/en/products/professional-graphics/amd-radeon-pro-w6400

So I am using nvidia a2000 in mine and it is 2.713" H x 6.6" and looks like the PRO W6400 is same size so should fit. TGP of 50W so your good on power.

I would argue that a Radeon is the go to confirmed working GPU though. Granted a2000/a4000 crowd doesn't have a heatsink/fan to fit inside the case yet but they work with bigger power adapters. Some people are not having any issues with nvidia cards and the stock power adapter.
 

Fazio

New Member
Dec 20, 2022
28
21
3
I would contribute the 15C more to a cold boot after paste application then anything else. For a true test you really want to let the machine warm up then put a consistent load on your CPU stock vs mx6/LM. I doubt your improvement will be 15C with new paste usually it's around 5C.

The thing you have to realize is Thermal Grizzly Conductonaut Extreme conductivity of 79W/mK while MX6 conductivity of 7.5W/mK. That is a big difference and I don't want people on here thinking they are going to get 15C from premium paste.
After 3 hours, the temperature is stable at 46.5C with same load and environment. It’s still a >10°C improvement for sure.