MPG X870E CARBON WIFI Beta BIOS

[Svet]

Spoiler: Older Versions

>>> E7E49AMSI.1A1E <<<

>>> E7E49AMSI.1A1H <<<

>>> E7E49AMSI.1A21 <<<

>>> E7E49AMSI.1A22 <<<

>>> E7E49AMSI.1A24 <<<

>>> E7E49AMSI.1A25 <<<

>>> E7E49AMSI.1A26 <<<

>>> E7E49AMSI.1A28 <<<

>>> E7E49AMSI.1A29 <<<

>>> E7E49AMSI.1A31 <<<

>>> E7E49AMSI.1A32 <<<

>>> E7E49AMSI.1A33 <<<

>>> E7E49AMSI.1A34 <<<

>>> E7E49AMSI.1A41 <<<

>>> E7E49AMSI.1A44 <<<

Notes:

• * Unstable RAM OC
• * Hang 0D when saving BIOS settings or doing M-Flash

>>> E7E49AMSI.1A62 <<<

>>> E7E49AMSI.1A65 <<<

>>> E7E49AMSI.1A66 <<<

>>> E7E49AMSI.1A93 <<<

>>> E7E49AMSI.1A99 <<<

 

[marceluss19]

Thank you.

From my experience with previous bioses, sometimes values of various params are not properly written after you set them. CO you set - 15, and after reboot you can see -5, scalar you set 2x, after reboot is 10x etc… Little problems it can causing…. Interesting, it just me or wider problem.

 

[uxexes]

Having voltage issues with both CPU and DDR5 RAM. Looks like it's not just me — many others are facing the same problems too.

 

[Anaksagoras]

Hello! At last, we have a stable BIOS! 7E49v1A4
RAM
MCLK=UCLK,
FCLK=2133

CPU
Curve Optimizer
All Cores -25

 

[dlomond]

Updated to .1A41

In the next bios can we get load back up profile from hard drive not just USB. This would be nice. Thanks for your updates.

 

[Alyred]

Hello! At last, we have a stable BIOS! 7E49v1A4
RAM
MCLK=UCLK,
FCLK=2133

CPU
Curve Optimizer
All Cores -25

Just curious: are you able to pass CPU+FPU+Memory tests with those settings? I've had significant issues getting 6400/2133 stable for more than 10m on previous BIOSes.

 

[DarkOmenX]

Dear MSI Support Team,


I hope this message finds you well.


I am writing to formally report an ongoing issue with my MSI MPG X870E KARBON WIFI
, which I have been experiencing since purchasing the product. Despite multiple attempts at adjusting BIOS settings, including voltage configurations and overclocking parameters, the motherboard consistently fails to maintain stable voltage levels after system reboots.


The system appears stable initially — all settings are applied correctly, and stress tests in Windows complete without errors. However, after one or more reboots, the settings no longer function as intended, leading to inconsistent performance and instability.


Given the nature of this behavior, I suspect there may be a manufacturing defect or firmware-related inconsistency affecting the board’s voltage regulation functionality.


My questions are as follows:

• How do you intend to resolve these issues?
• Will you accept the motherboard as a defective unit and proceed with a replacement process?

I would greatly appreciate your guidance on the next steps, particularly regarding whether I should return the unit for inspection or if there is a potential BIOS update or workaround that could address this concern.


Looking forward to your prompt response.


Best regards,

Bios: A41

Firstly, this is not a support forum, so MSI likely won’t care about what you post here.
Secondly, if you attempt to RMA the board or suspect it is defective in any way, you should be using an officially released MSI BIOS, which A41 is not.
Before opening a ticket with MSI official support, roll back the BIOS to the latest official version and use the BIOS default settings.
Even better, perform a full CMOS clear and then load the BIOS defaults.

 

[DarkOmenX]

Just curious: are you able to pass CPU+FPU+Memory tests with those settings? I've had significant issues getting 6400/2133 stable for more than 10m on previous BIOSes.

I have stable 6400/2200 MLCK=UCLK but it was also stable on latest official bios 7E49v1A4 so that must be more due to RAM capability then board.
Ram is Corsair Titanium 7200 48 GB CL 36 1.4V

 

[sebastian.eberhar158d02e0]

With 7E49v1A4 I get a freeze after the boot logo is shown after a cold start, AGAIN.
I am so close to buying a new board from another brand. Not sure if they are as bad, but it seems worth the try.

 

[Anaksagoras]

Hey, I did some tests yesterday and shared the screenshots. I played Diablo 4, and it was smooth sailing. What kind of RAM tests are you talking about? I can run tests on those too.

 

[DarkOmenX]

With 7E49v1A4 I get a freeze after the boot logo is shown after a cold start, AGAIN.
I am so close to buying a new board from another brand. Not sure if they are as bad, but it seems worth the try.

I had a similar experience before with the X570ACE and X570S boards, which drove me nuts despite others swearing everything worked perfectly on their end. Frustrated with all the instability issues, I made an epic move and switched to the even worse 13900K, followed by another potential disaster, the 14900 K. Ironically, I am now on the opposite side with the 9800X3d+Carbon WiFi, which is working flawlessly so far . I think that's what is called "The AMD Thing." You never know what you're gonna get.
Intel is way more predictable and stable in this regard, however, they fckd it up badly with 13/14 gen .

 

[Alyred]

I have stable 6400/2200 MLCK=UCLK but it was also stable on latest official bios 7E49v1A4 so that must be more due to RAM capability then board.
Ram is Corsair Titanium 7200 48 GB CL 36 1.4V

I have both Corsair Titanium 8000 48G and 6400 64G (currently running the 6400) but would almost immediately bluescreen once Windows loaded when running at 6400 M=U. I haven't tried pushing that in quite a while though, so maybe I'll give it another shot with the newer BIOS. Been since around A25 that I tried pushing it like that.

 

[Alyred]

I have stable 6400/2200 MLCK=UCLK but it was also stable on latest official bios 7E49v1A4 so that must be more due to RAM capability then board.
Ram is Corsair Titanium 7200 48 GB CL 36 1.4V

I have both Corsair Titanium 8000 48G and 6400 64G (currently running the 6400) but would almost immediately bluescreen once Windows loaded when running at 6400 M=U. I haven't tried pushing that in quite a while though, so maybe I'll give it another shot with the newer BIOS. Been since around A25 since I gave it a shot and newer BIOSes are probably more stable.

Hey, I did some tests yesterday and shared the screenshots. I played Diablo 4, and it was smooth sailing. What kind of RAM tests are you talking about? I can run tests on those too.

I run for stability, so I typically run Aida64 burn-in with CPU, FPU, and Memory selected. I usually call it good if it runs for 12 hours without an issue. I've seen it fail at 8 hours before, but that's probably still stable for 99% of folks since the CPU and memory just aren't usually pushed that hard in normal use. Other folks use ycruncher but I haven't really learned that one yet. OCCT seems to suss out cpu undervolting stability fairly well too.l, but doesn't appear to push memory quite as hard.

 

[DarkOmenX]

I have both Corsair Titanium 8000 48G and 6400 64G (currently running the 6400) but would almost immediately bluescreen once Windows loaded when running at 6400 M=U. I haven't tried pushing that in quite a while though, so maybe I'll give it another shot with the newer BIOS. Been since around A25 since I gave it a shot and newer BIOSes are probably more stable.

I run for stability, so I typically run Aida64 burn-in with CPU, FPU, and Memory selected. I usually call it good if it runs for 12 hours without an issue. I've seen it fail at 8 hours before, but that's probably still stable for 99% of folks since the CPU and memory just aren't usually pushed that hard in normal use. Other folks use ycruncher but I haven't really learned that one yet. OCCT seems to suss out cpu undervolting stability fairly well too.l, but doesn't appear to push memory quite as hard.

While my Titanium 7200 works nicely with 6400/2200 and no BSOD or errors during a 30-minute run of AIDA stress test CPU+FPU or OCCT, I can't get it to work on 8000 at all. Sometimes it will go into Windows, but a restart or shutdown and power on ends up with a black screen and a distinctive memory not recognized beep. I have the buzzer on my board.

 

[Anaksagoras]

I run for stability, so I typically run Aida64 burn-in with CPU, FPU, and Memory selected. I usually call it good if it runs for 12 hours without an issue. I've seen it fail at 8 hours before, but that's probably still stable for 99% of folks since the CPU and memory just aren't usually pushed that hard in normal use. Other folks use ycruncher but I haven't really learned that one yet. OCCT seems to suss out cpu undervolting stability fairly well too.l, but doesn't appear to push memory quite as hard.

I totally get it, forcing the system for hours and searching for errors can make anyone feel a bit paranoid. I did have a few teething problems with the previous BIOSES, but I'm happy to say that I'm now fully satisfied with the A4 BIOS. I know it's not necessary for me to spend hours doing tests. I just want to be able to use my PC without any problems, and that's all that matters to me.

 

[dlomond]

I was on 41 but temps seems to be about 5 to 8 degrees hotter I have to do a little more testing . I went back to 40 and wil test. I just trying out all the beta bios but I don't have any problem with this system so it's hard to find problems when I don't have any.

 

[Anaksagoras]

Bios 7E49v1 A4
RAM
MCLK=UCLK,
FCLK=2133
CPU
Curve Optimizer
All Cores -30
+200

​

 

[Bizkitca156902da]

Hello! At last, we have a stable BIOS! 7E49v1A4
RAM
MCLK=UCLK,
FCLK=2133

CPU
Curve Optimizer
All Cores -25

are you on Realtek PCIE Network Drivers 10.73 or 10.74?
I am debating upgrading BIOS (7E49v1A1 2024-09-05) -> (7E49v1A4 2025-05-16) due to new RAM coming in tomorrow... which I will try on the November BIOS first? Last time I tried to upgrader BIOS it would not post, on 3 different versions, so I have reverted to and stayed on 1A1

which between when I ordered it last week and now is already EOL...?!

Was only on the market for a few months and is already EOL... what / why?

 

[Alyred]

are you on Realtek PCIE Network Drivers 10.73 or 10.74?
I am debating upgrading BIOS (7E49v1A1 2024-09-05) -> (7E49v1A4 2025-05-16) due to new RAM coming in tomorrow... which I will try on the November BIOS first? Last time I tried to upgrader BIOS it would not post, on 3 different versions, so I have reverted to and stayed on 1A1

which between when I ordered it last week and now is already EOL...?!

Was only on the market for a few months and is already EOL... what / why?

They've made absolutely massive improvements in memory stability between 1A1 and 1A4. I would definitely upgrade.
Many of the BIOS upgrades have had trouble POSTing after the upgrade, make sure to do a BIOS RESET flash after you've upgraded and it's rebooted. Also, some of the BIOS updates in the meantime have an absolutely brutal memory training time, upwards of 7 minutes. After the upgrade, wait for it to restart and give it some time to make sure the BIOS update has finished, then shut down and do a full 20 second BIOS reset by holding the button on the back of the motherboard with the power supply switched off. This has solved a lot of BIOS upgrade problems for me with this board, specifically.
If you get to code 15, that's training the RAM. Give it 10 minutes before giving up. It may take a while, then turn on Memory Context Restore in the BIOS once you get in.
With memory that tight of timing, it may take even longer to train, or may not do so at all at XMP settings - the IMC on these 9800x3ds seems to be very finicky. You may need to manually tune the RAM to run it outside of JDEC.
As for the EOL - I've heard g.skill has had a significantly difficult time binning those chips in enough quantity to meet demand and may be dropping the CL26 memory production for the time being, and once the supply runs out, there won't be any more for a while.

 

[Alyred]

Bios 7E49v1 A4
RAM
MCLK=UCLK,
FCLK=2133
CPU
Curve Optimizer
All Cores -30
+200

​

Those are some good numbers, but Cinebench isn't really a stability test, it's a raw performance test. Good way to warm up the CPU and see your temperature curves for sure. What scalar are you running with (if you manually set it)?

If you can run AIDA64 Burn-in with CPU, FPU, cache, and System Memory all being tested for 15-20m, you've probably got your settings stable enough for 99% of any gaming or workload. If not, I'd dial the CO back by 5 or so. Playing with the curve shaper can also help stabilize a system at higher load, but @vic123159d02e5 is more knowledgeable on that than I am.

Looking at your ZenTimings screenshot and Aida64 numbers, you might be able to cut your latency significantly by lowering your tRFC to 500 and raising your tREFI to 50000, just for some easy settings. Even the EXPO/XMP timings on those two settings are generally WAY looser than they need to be, particularly since those are Hynix A-die so far as I'm aware. You're not running the CPU in Hypervisor mode (according to your AIDA screenshot) so you should be getting ~60ns timings. thought with the IMC on the 9950x3d there might be differences from the 9800x3d that I'm not aware of. Guess I better learn fast since I have a 9950x3d coming Friday...

 

[Anaksagoras]

are you on Realtek PCIE Network Drivers 10.73 or 10.74?
I am debating upgrading BIOS (7E49v1A1 2024-09-05) -> (7E49v1A4 2025-05-16) due to new RAM coming in tomorrow... which I will try on the November BIOS first? Last time I tried to upgrader BIOS it would not post, on 3 different versions, so I have reverted to and stayed on 1A1

which between when I ordered it last week and now is already EOL...?!

Was only on the market for a few months and is already EOL... what / why?

Realtek PCIE Network Drivers 10.73.815.2024

A4 is definitely the most stable bios.