PCIe bug with M2_1 slot on X870E Tomahawk

Status
Not open for further replies.

jlsmith.01712b902af

New member
Joined
Feb 16, 2025
Messages
15
So, the M.2_1 slot keeps reverting to PCIe 1 x 4 on a cold boot, sometimes just a reboot from within Windows. Manually setting it to Gen 4 or 5 within the BIOS will get it to run full speed until the next shut down/cold boot, then it's back to 1 x 4 again. Basically, I have to keep switching between gen 4 and 5 setting on every boot or it's stuck at 1 x 4. Auto is also broken. Samsung 990 Evo Plus 1TB.

I've seen another user on Reddit with the exact same issue. Same board, different drive. Anyone using this board, check your drive speed!
 
This is bullshit. You can be sure that a RMA will be useless: you will receive the exact same motherboard with the exact same problems. Since BIOS A41 my Tomahawk M2_1 works perfectly (I don't use RAID). Wich means it can be fixed via BIOS. I don't see how it would be any different for the Godlike.

This is just classic MSI. Godlike is $1000, Tomahawk $300. Because of the price alone, I imagine that Tomahawk users far outnumber those of the Godlike. It's just a lot easier for them to try and bury the issue for the Godlike than for the Tomahawk, that's all.
For me the "fix" with 2A41 BIOS is not 100% effective. In my case, I installed 2A41 on March 18 and everything was perfect - daily testing and monitoring, zero issues. Then on March 27 I woke my system from a sleep state and logged my first instance where Samsung Magician and CrystalDiskInfo both reported "1x4" transfer mode, but CrystalDiskMark tested with 4x4 speeds. After numerous tests I was also able to record 1x4/3x4 results all while showing "1x4" as the transfer mode. After several tests, reboots and wakes, I decided to clear the CMOS. Afterwards, I was back to 4x4/4x4 for almost another two weeks. This past Saturday night, I woke from sleep to find I was showing "4x4" transfer mode in both Magician and CrystalDiskInfo, but then benched at 1x4 performance in both Magician and CrystalDiskMark. After a cold boot, things are back to normal again and I have not encountered another mismatch between the reported transfer rate and benchmark testing, nor have I had to clear my CMOS again.

At least prior to 2A41, I was getting parity between the reported transfer mode and the benchmark performance - it was almost always the wrong one, mind you, but at least it was consistent. Now I feel like I can't really trust the reported transfer mode no matter what tool I'm looking at. Going almost 2 weeks in between occurrences is a HUGE improvement, don't get me wrong. Something is mostly working with 2A41, but it's still a pesky issue and I know of at least two other users on reddit that have similar experiences. Ultimately, the only result I can trust is the benchmark tests. Given my experiences so far, I suspect that whatever tool you use to check the transfer mode could be misleading and the only way to know for sure is to run the bench tests and see how it truly performs.
 
Here is the reply I received from MSI Support:

Ticket:
SSD in M.2_1 socket detecting as PCIe Gen. 4x1

Content:
At this point it seems that RMA is going to be necessary as I have not heard any updates of any fix being possible over bios update. You can request an RMA here: https://service.msicomputer.com/msi_user4/support/rma_form.aspx

-=-=-=-

I'm not convinced that if I go through the hassle of removing this board, dealing with the downtime while waiting for a replacement, that I'll get a new board without this problem. A $1000 motherboard and this issue gets out of QA. This is the first MSI board I have bought, and will be the last.
 
Here is the reply I received from MSI Support:

Ticket:
SSD in M.2_1 socket detecting as PCIe Gen. 4x1

Content:
At this point it seems that RMA is going to be necessary as I have not heard any updates of any fix being possible over bios update. You can request an RMA here: https://service.msicomputer.com/msi_user4/support/rma_form.aspx

-=-=-=-

I'm not convinced that if I go through the hassle of removing this board, dealing with the downtime while waiting for a replacement, that I'll get a new board without this problem. A $1000 motherboard and this issue gets out of QA. This is the first MSI board I have bought, and will be the last.
That context helps. "i have not heard of any fix being possible over bios update" is very different than "we found the issue and it required a physical change to the board to fix it" this means if you get a replacement it will still be broken. An RMA will not fix anything.
 
I had opened an incident ticket earlier today for my X870E Tomahawk WiFi on the m2_1 issue
They got back to me after a couple of hours later and shared a new BIOS (E7E59AMSI.2A41). This has restored my PCIe 5.0 x4 speed and maintained it after several cold/warm restarts.
I believe when I was doing the reflash, the description mentioned BETA. Going into the setup now, it only lists the BIOS version as above.
Not sure if this version is any different from the 7E59v2A5 that was up last week and pulled down.
 
Here is the reply I received from MSI Support:

Ticket:
SSD in M.2_1 socket detecting as PCIe Gen. 4x1

Content:
At this point it seems that RMA is going to be necessary as I have not heard any updates of any fix being possible over bios update. You can request an RMA here: https://service.msicomputer.com/msi_user4/support/rma_form.aspx

-=-=-=-

I'm not convinced that if I go through the hassle of removing this board, dealing with the downtime while waiting for a replacement, that I'll get a new board without this problem. A $1000 motherboard and this issue gets out of QA. This is the first MSI board I have bought, and will be the last.
I had opened an incident ticket earlier today for my X870E Tomahawk WiFi on the m2_1 issue
They got back to me after a couple of hours later and shared a new BIOS (E7E59AMSI.2A41). This has restored my PCIe 5.0 x4 speed and maintained it after several cold/warm restarts.
I believe when I was doing the reflash, the description mentioned BETA. Going into the setup now, it only lists the BIOS version as above.
Not sure if this version is any different from the 7E59v2A5 that was up last week and pulled down.
If your using Samsung or Western Digital try a different Make of SSD if possible
 
Here is the reply I received from MSI Support:

Ticket:
SSD in M.2_1 socket detecting as PCIe Gen. 4x1

Content:
At this point it seems that RMA is going to be necessary as I have not heard any updates of any fix being possible over bios update. You can request an RMA here: https://service.msicomputer.com/msi_user4/support/rma_form.aspx

-=-=-=-

I'm not convinced that if I go through the hassle of removing this board, dealing with the downtime while waiting for a replacement, that I'll get a new board without this problem. A $1000 motherboard and this issue gets out of QA. This is the first MSI board I have bought, and will be the last.
"SSD in M.2_1 socket detecting as PCIe Gen. 4x1"

Did you give them the right information? The drive is being detected as Gen 1 x4 NOT Gen 4 x1
 
I just got a response in my ticket stating that the bug cannot be fixed via BIOS and the board must be RMA'd (at least for the Godlike)
What exactly was your issue with your Godlike. I'm trying to compare.

My issue is that I cannot soft reboot without the OS boot process taking forever and the mouse stuttering. Forcing the system to power down and power back on corrects the issue. Running A30 for my X870e Godlike. Giant waste of money. This board is a complete failure. Why is there no beta BIOS for the top tier board at $1600?

I'm using only the m2_1 slot. Does MSI have a tool to help me send Diagnostics logs?

Also my Asus Astral 5090 OC LC does not work on this board properly. The second HDMI simply does not work. Moving the card to a different motherboard fixes the issue.
 
Last edited:
What exactly was your issue with your Godlike. I'm trying to compare.

My issue is that I cannot soft reboot without the OS boot process taking forever and the mouse stuttering. Forcing the system to power down and power back on corrects the issue. Running A30 for my X870e Godlike. Giant waste of money. This board is a complete failure. Why is there no beta BIOS for the top tier board at $1600?

I'm using only the m2_1 slot. Does MSI have a tool to help me send Diagnostics logs?

Also my Asus Astral 5090 OC LC does not work on this board properly. The second HDMI simply does not work. Moving the card to a different motherboard fixes the issue.
The only issue I have is the M2_1 drive speed issue. I've switched to M2_2 as a workaround.
I've not had the boot issue you mentioned.
 
I had opened an incident ticket earlier today for my X870E Tomahawk WiFi on the m2_1 issue
They got back to me after a couple of hours later and shared a new BIOS (E7E59AMSI.2A41). This has restored my PCIe 5.0 x4 speed and maintained it after several cold/warm restarts.
I believe when I was doing the reflash, the description mentioned BETA. Going into the setup now, it only lists the BIOS version as above.
Not sure if this version is any different from the 7E59v2A5 that was up last week and pulled down.
Can you share that BIOS please?!
I have the same MB and same problem.
Cheers!
 
So, the M.2_1 slot keeps reverting to PCIe 1 x 4 on a cold boot, sometimes just a reboot from within Windows. Manually setting it to Gen 4 or 5 within the BIOS will get it to run full speed until the next shut down/cold boot, then it's back to 1 x 4 again. Basically, I have to keep switching between gen 4 and 5 setting on every boot or it's stuck at 1 x 4. Auto is also broken. Samsung 990 Evo Plus 1TB.

I've seen another user on Reddit with the exact same issue. Same board, different drive. Anyone using this board, check your drive speed!
I had the problem of "lottery PCIe boot" with the last official Bios 7E59v2A4 and my Crucial T705 SSD. Now, with the same bios, having changed the SSD to a Samsung 9100 Pro, I constantly boot in PCIe 5, either in cold or warm boot.
 
I had the problem of "lottery PCIe boot" with the last official Bios 7E59v2A4 and my Crucial T705 SSD. Now, with the same bios, having changed the SSD to a Samsung 9100 Pro, I constantly boot in PCIe 5, either in cold or warm boot.
That's interesting as a lot of the posters here were using Gen 4 drives (Samsung 990 mostly).
Is your M2_1 slot set to Auto or Gen 5?

The 9100 is newer than all the other Gen 5 drives and uses Samsung's own controller

No guarantee it will keep working at Gen 5, it could do 100 correct then fail
 
The M2_1 issue has nothing to do with the kind of CPU. And for those having the issue with the x870e Tomahawk, try the beta bios 2A41 (link 6 messages above). It fixes the issue for most people.
 
Last edited:
Status
Not open for further replies.
Back
Top