PCIe bug with M2_1 slot on X870E Tomahawk

J

jlsmith.01712b902af

New member
Joined
Feb 16, 2025
Messages
14
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!
 
Sort by date Sort by votes
Just had this from support.

“7E59v2A41 is newer than 7E59v2A4 which is the one on MSI website now), please keep using 7E59v2A41 BIOS becasue the solution is added in this version” Thank you!

Previously they said the beta is no different to the public one now on the website.

This might explain why I’m currently not getting the bug atm and those who are using the public new version are still getting the bug.
 
Upvote 0 Downvote
I have been running the beta BIOS now for a bit more than 24 hours. Here's a summary:

--Prior to the beta my X870E Tomahawk was 100% stuck on 1x4 speeds with my Samsung 990 Pro 2TB in slot M.2_1. No random changes or fluctuations between cold boots and restarts. Nothing in the BIOS affected performance (eg, toggle Gen4/Gen5 in BIOS, power management options, etc.). 24/7, the drive would only show 1x4 and CrystalDiskMark was clearly showing 1x4 performance levels. PC build date was 2/14. Board produced in January 2025. BIOS versions 7E59v2A2 (initial), 7E59v2A3, and 7E59v2A33 all had the same M.2_1 issue.​

--Yesterday, 3/17, I just went and grabbed the beta from a Google drive link posted elsewhere after getting frustrated with support (more on that below). After the initial flash, nothing changed so then I shutdown completely, unplugged the PSU and cleared the CMOS. Upon cold booting after CMOS was cleared, I let it go straight to Windows where I was able to see that my Samsung 990 Pro 2TB in slot M.2_1 was functioning at the correct speed for the very first time and CrystalDiskMark test results were within spec of 4x4 performance.​
​
--I then went to restart via Windows so I could re-enter BIOS to enable EXPO, but upon restarting via Windows I encountered a "C5" error on the EZ display before the system could cycle off. I have never encountered C5 before or since. I've had the occassional "0d" on Windows restart, but I think that's a topic for a different discussion (a separate board issue relating somehow to performance monitoring software). After after about 5 minutes of hanging at the C5 error I had to force power shutdown on the case.​
​
--Lastly, I powered the PC back up, accessed BIOS without any issues, re-enabled EXPO and booted to Windows. Again, M.2_1 was registering 4x4 and CrystalDiskMark measured as expected.​
​
Since then, I've had two more full shutdown/cold boot cycles and one restart. I've conducted tests each time. No issues with either the M.2_1 performance nor any other shutdown/memory issues (C5 or 0d). I will continue to monitor and document any changes, but for now I am much more pleased.

I will say, however, that none of this is thanks to MSI support going back to 2/25. My ticket thread is a log of poor-to-insulting customer service antics, copy/paste support handbook responses and, in the case of yesterday, outright misinformation (I was told that there was no such thing as a beta BIOS - and this was in direct response to me showing them the Google drive link)

The X870E Tomahawk is a very capable and decent board considering it's low price point compared to other X870E boards out there, but I do sometimes feel like I got what I paid for. I guess I should have taken the wrong wifi antenna pack-in as a warning sign when I opened the box (they had a batch go out with white antennas that completely did not match the board). At least support was capable enough to send me a black one quickly even if it wasn't exactly the correct model antenna shown in the marketing materials.

Edit: just to respond to Krisk7157302df's post above mine. I am still running the completely non-existent beta 7E59v2A41.
 
Last edited:
Upvote 0 Downvote
After i flashed the beta bios from the link above (Thank you) the bug is gone. But didn't stuck at C5 or any issues while booted directly into bios setup and changed "Installer -> Disabled, Profile 1 and set DRAM at 6400 MHZ". Saved and booted to Windows without any issues.

CrystalDiskMark speeds are normal now.
 
Upvote 0 Downvote
rvillatoba14ec02d0 said:
Have you tried the beta bios, since the official one doesn't work in your case? I know they are supposed to be the same, but I also frequent the msi forum on Reddit and people seem to have different results depending on the bios (new one or beta).
Click to expand...
2A41 is the beta but 2A4 is the one on the support page?
There will be some slight differences other than the BIOS version number I imagine.
 
Upvote 0 Downvote
I would give it a week at least.
My Godlike went 5 days without the problem before it showed again. I still get the issue as there's no new BIOS yet.
Sometimes I can go a few days, sometimes it'll show up twice in a row
 
Upvote 0 Downvote
Has anyone with a Godlike board got any responses from support to indicate they acknowledge this issue ALSO exists on the Godlike? Normally I would take a fix for the Tomahawk as a good indicator they will fix all the other boards too, but after failing to replicate the issue for so long (wtf were they even doing??), deflecting blame to the ssd maker, denying the problem exists at all, and letting this problem persist for TWO generations of motherboards now. I... do no have a ton of faith in them anymore. And I am still concerned they will write this one off as done and move on if we dont keep hounding them to also fix it on the Godlike boards.
 
Upvote 0 Downvote
This is actually good news, then; the beta version seems to have fixed the bug. I'm rather relieved, although we shouldn't declare victory too quickly.
I'm fairly confident the fix will be applied to all MSI X870e boards, including Godlike, once it becomes official.
 
Upvote 0 Downvote
daniel.mcgowan1160b02f1 said:
Has anyone with a Godlike board got any responses from support to indicate they acknowledge this issue ALSO exists on the Godlike? Normally I would take a fix for the Tomahawk as a good indicator they will fix all the other boards too, but after failing to replicate the issue for so long (wtf were they even doing??), deflecting blame to the ssd maker, denying the problem exists at all, and letting this problem persist for TWO generations of motherboards now. I... do no have a ton of faith in them anymore. And I am still concerned they will write this one off as done and move on if we dont keep hounding them to also fix it on the Godlike boards.
Click to expand...
Still testing it was the last response I had a few days ago.
I'll be replying to their support email every week asking for a status report
 
Upvote 0 Downvote
On my MAG X870E TOMAHAWK WIFI the latest official BIOS does not enable the full speed of NVMe. As far as the beta version posted by a user in this forum, that one too has issues. Works randomly. So, yeah. No go. What is crazy is, this is basic stuff that should of been addressed from the get go as standard requirement. It's not like a new future that they would need to take the bugs out because is so new. In addition, i am using RAID 1, so what i have is, one drive stays steady at Gen4, but the M2_1 fluctuates on beta version for the very first time from Gen3 to Gen4. When M2_1 is on Gen3 i get 1,6Gb read and 850Mb write, once i do several restarts and somehow it gets on Gen4 the M2_1 slot, it goes up to 10,5Gb read and 8Gb write. So yeah, it might work on standalone NVMe cards but in RAID mode or better yet RAID1 that i use, this is still an issue.
 
Upvote 0 Downvote
The beta A41 is different to the public one on the downloads page. After 3 days of extensive testing of beta A41 on my Tomahawk X870E, I’m considering this issued now fixed, as nearly 60 restarts, 45 shutdowns (boot from cold) and 35 wake up from sleep, the drop to 1.0 x4 has not occurred and neither has benchmarking speeds dropped (all shown what is expected from the NVME)

I don’t plan to do any more tests, as previously on the other bios version it would have occurred at least 10 times,
if I did the same amount of restarts, etc, and other users who were getting it every power on/restart have reported it gone on the beta A41.

For RAID, I cannot comment because I don’t use it.
 
Upvote 0 Downvote
This should sum up what i am seeing under RAID1:
NVMe-speed-results-gen3-gen4-mix.png


Also, does anyone know as to why in Advanced mode under the PCI Express settings when opening the dropdown menu from Auto it can only show up till Gen4 not Gen5 as the rest of the PCIe ports in the drop down menu? I thought x870e is PCIe 5.0, so why only allows me to go up to Gen4? I find that one odd, and again, it does not make any sense whatsoever.
 
Last edited:
Upvote 0 Downvote
I can confirm that I moved my Samsung 990 Pro from M2_2 to the M2_1 slot and manually set M2_1 to Gen4, but the disk is still running at PCI Gen 1x4 speed. I have installed the latest firmware 7E59v2A4.

Snímek obrazovky 2025-03-19 202041.png
 
Upvote 0 Downvote
Back
Top