MPG X870E CARBON WIFI Beta BIOS

Svet

Well-known member
Administrator
Joined
Nov 20, 2003
Messages
85,347
>>> 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 <<<
 
Last edited by a moderator:
Well, in that case, sounds like a bug related maybe to BIOS (maybe AGESA) and CPU-Z version. On my old 7950X3D, lowering IF to 2133MHz from 2167MHz did the trick and CPU-Z would load without issues. At 2167 and above, it would hang at 70%, even after voltage fine-tuning on vSOC, VDDGs, VDDP.

No issues with 9950X3D at IF 2200MHz on A28.
 
Хорошо, тогда все хорошо. Странно, у меня была та же проблема, которая случилась ровно при 70% загрузке в CPU-Z, но я сильно разогнал ПЧ и скорость памяти (немного слишком сильно ;) ).
Unfortunately, it's the same even with default settings.
Now I’m starting to stress that my CPU or motherboard might be defective.
 
The offical version number 7E49v1A33 is published on the download page. Here is the full changelog:

- AGESA PI-1.2.0.3c released.

Is that joke?
 
The offical version number 7E49v1A33 is published on the download page. Here is the full changelog:

- AGESA PI-1.2.0.3c released.

Is that joke?
Is this the same as the one published by SVET? I have quite good experience with the one from SVET A33
 
Is this the same as the one published by SVET?

Yes it is.

Hey Svet, is there any way you can let us know what fixes or changes are in these beta BIOS releases? It would help all of us target those fixes in testing and we could provide you valuable feeback if things are working better etc. We know these are beta and should use them at our own risk.. so for those of us willing to try them, please help us help you!

Nope, no info about those sorry. Can't provide such info.
 
Unfortunately, it's the same even with default settings.
Now I’m starting to stress that my CPU or motherboard might be defective.
CPU-Z version 2.15 did you try that one (that's the one I'm using) ? 2.13 is also known to be ok .
I do not know is that same version. But it's a bit strange that it only comes with the AGESA update when there are so many things that need to be fixed.
The A32 already came with that fix, so I feel like they're not telling us everything. It seems like they're trying to address certain issues, but they can't really endorse any changes unless they're confident those changes work. We're essentially the guinea pigs here. That said, the A33 seems pretty solid. To be honest, I haven't exactly been kind to MSI in the past, but I think they've done a great job with this one, until ... I say the opposite ;)
To add information from my previous test of the A33, it seems that, for CPPC parameters, the driver value provides the best stability and delivers great overall numbers in benchmarks.
EDIT : For my latest testing tonight, I managed to regain single-thread performance with a 104.5 BCLK(only trade off, was setting my DDR5 speed lower, it seems no matter what voltage I throw in, it would not accept more than 8360 mhz). These sessions weren’t particularly long, so I can't guarantee it’s 100% stable yet. However, it successfully passed tests using OCCT, Cinebench R23, AIDA64, CPU-Z, and Y-cruncher (which showed the most impressive improvement), even 3dmark is showing me some love in the numbers it displays ;) .
PS: disturbing how low I can set CO curve, -46 right now (-44 was maximum with A32 and -36-37 with A25)
 
Last edited:
I'm sorry if all or part of what I'm about to say is something you already know, but I'm not sure what your current overclocking level is.
It definitely sounds like your GPU doesn't like the increased PCIe frequency, same thing happened with my 4090. Unfortunately, aside from dialing back the BCLK, there’s not much you can do.
Also :
- Keep in mind that when overclocking the 9800X3D, you'll probably hit around 102–102.5 BCLK before you see diminishing returns in multi-core performance. Your single-thread performance should still improve though, which is usually what you’re aiming for in gaming.
- Good cooling is key to a stable overclock, so if you don't have a solid AIO or custom watercooling, I wouldn't recommend going beyond standard PBO +200 MHz.

There are a few BIOS settings that can help your overclock, along with some software tools you'll need to monitor and test your speeds. This isn’t a complete list, I’m at work right now and not in front of my home PC, so I might miss a few.

BIOS:
  • Disable Integrated Graphics
  • Advanced / Integrated Graphics Configuration
  • Disable Spread Spectrum
  • OC / Advanced CPU Configuration
  • Set Limit to "Advanced" / Motherboard (removes power limit)
    Scalar: Manual, x10 (gives more headroom for overclocking)
  • OC / Advanced CPU Configuration / PBO
  • Load-Line Calibration:
    For CPU, 8 gives max overclocking room, but 3–5 is usually most stable
    For VSOC, I recommend 3
    CPU Switching Frequency: 1000 kHz (faster response, better for stability when overclocking)
    CPU Vcore Overvoltage Protection: 400 mV(gives headroom for spikes during heavy load, useful in OC)
    VSOC Switching Frequency: 1000 kHz
  • OC / near bottom in DigiALL POWER
  • Set UCLK = MEMCLK (on DDR5 6000–6400 configs)
  • OC / UCLK DIV1 MODE
    Set Infinity Fabric to 2000 MHz (keep in mind it scales with BCLK , for example, at 103.5 BCLK, 2000 , becomes 2070 MHz)
    • OC /FCLK Frequency

    • SOFTWARE :
Monitoring, nice and easy to use, mostly accurate.
HWINFO64

Monitoring, bench and memory test ( Actually I only use it to test memory stability and run memory benchmark, because its accuracy is really bad at monitoring and other benches IMO)
AIDA64

Great stability tester for memory, for quick test use option 0, 1, 8 (great for a quick test when running 32 GB memory)
Y-Cruncher

Quick multi-thread and single thread Benchmark ( not the most reliable but the one I use almost on every boot to see if my speeds are "normal" and actually the single thread bench is not that bad)
CPU-Z

Good mult-thread benchmark (particularly useful to see if your CO curve is stable !) , Ok single thread benchmark but a bit too slow IMO.
CinebenchR23

Great Cpu stability tester ( but really poor benchmark value IMO)
OCCT

I may add a few tweak tonight when i get home.
One more time, thank you for the guidelines
Using these guidelines and the latest V33 Bios, I was able to tune my PC to my desires. It works smoothly and remains stable.
blck 102.5 on msi 4090 Liquid X (no black screen)
The rest as per the guidelines, now, memory, I was kinda fighting this.The Corsair TITANIUM 7200 works In a wide range of frequencies and timings, but somehow I find the best result being on XMP + a bit of OC over the blck102.5 The memory has 2 A-XMP profiles, one on 7200 and the second on 7400; however, it sits between now.
Currently, the fabric is set to 2000 MHz, and I'm unsure if I should adjust this setting or what benefits might come from doing so.
I've seen various opinions on the matter.
In comparing AMD and Intel systems, I find that tuning AMD can be quite challenging. A small mistake often requires a BIOS reset, while with Intel, not so much. then once I found a stable configuration, I could usually apply it after each BIOS update without issues.
AMD, on the other hand, feels more unpredictable, as settings that work well now may not remain stable after a BIOS update, leading to a lot of "chasing" for optimal performance.I have some past experience with AMD, and I wasn’t overly impressed at the time.
However, after the numerous issues Intel produced with 13th/14th gen, I decided to give AMD another chance.
This time, it seems significantly improved, more stable, and the performance of the 8-core CPU, the 9800X3D, is truly impressive.
I’m thankful to SVET for sharing the BETA BIOS for those who want to experiment with it, despite the risks involved. However, I’m curious why MSI has removed the BETA BIOS from the regular motherboard download page?
Occasionally, they seem to post it, but I get the impression that it is now primarily accessible to a smaller group of enthusiasts and forum administrators.
 
One more time, thank you for the guidelines
Using these guidelines and the latest V33 Bios, I was able to tune my PC to my desires. It works smoothly and remains stable.
blck 102.5 on msi 4090 Liquid X (no black screen)
The rest as per the guidelines, now, memory, I was kinda fighting this.The Corsair TITANIUM 7200 works In a wide range of frequencies and timings, but somehow I find the best result being on XMP + a bit of OC over the blck102.5 The memory has 2 A-XMP profiles, one on 7200 and the second on 7400; however, it sits between now.
Currently, the fabric is set to 2000 MHz, and I'm unsure if I should adjust this setting or what benefits might come from doing so.
I've seen various opinions on the matter.
In comparing AMD and Intel systems, I find that tuning AMD can be quite challenging. A small mistake often requires a BIOS reset, while with Intel, not so much. then once I found a stable configuration, I could usually apply it after each BIOS update without issues.
AMD, on the other hand, feels more unpredictable, as settings that work well now may not remain stable after a BIOS update, leading to a lot of "chasing" for optimal performance.I have some past experience with AMD, and I wasn’t overly impressed at the time.
However, after the numerous issues Intel produced with 13th/14th gen, I decided to give AMD another chance.
This time, it seems significantly improved, more stable, and the performance of the 8-core CPU, the 9800X3D, is truly impressive.
I’m thankful to SVET for sharing the BETA BIOS for those who want to experiment with it, despite the risks involved. However, I’m curious why MSI has removed the BETA BIOS from the regular motherboard download page?
Occasionally, they seem to post it, but I get the impression that it is now primarily accessible to a smaller group of enthusiasts and forum administrators.
I'm also much more experienced with Intel CPUs, but I’ve spent an enormous amount of time these past months tweaking the 9800X3D, my fourth AMD CPU. Before that, I think I owned around 12–14 Intel CPUs. (Yes, I'm old enough for that! 😉) My first PC was hosting an Intel 8088 processor (IBM PC), running at 5 or 8 MHz if I remember correctly, and I was about 10 years old at the time.
To answer your question, a higher Infinity Fabric frequency would typically enhance both single-thread and multi-thread performance. However, keep in mind that if it doesn’t work, you might have to deal with another BIOS reset. 😉
Also, don’t forget that IF scales with your BCLK. For example, at 2000 IF, you’re already running around 2050 MHz. If I were you, I’d try increasing it step by step and benchmarking as you go. In my case, at 104.5 BCLK, I’m running 2067 IF, which sits at approximately 2158 MHz. I attempted pushing it higher, but unfortunately didn’t have any luck. For a day to day use, I think I'll keep it @104 bclk
I think I'm just about done tweaking the CPU on this board. I'm just wondering how far it could have gone with an Eclk board...tried 105bclk, but Y-cruncher wouldn't pass nor cinebench r23.
 
Last edited:
I'm also much more experienced with Intel CPUs, but I’ve spent an enormous amount of time these past months tweaking the 9800X3D, my fourth AMD CPU. Before that, I think I owned around 12–14 Intel CPUs. (Yes, I'm old enough for that! 😉) My first PC was hosting an Intel 8088 processor (IBM PC), running at 5 or 8 MHz if I remember correctly, and I was about 10 years old at the time.
To answer your question, a higher Infinity Fabric frequency would typically enhance both single-thread and multi-thread performance. However, keep in mind that if it doesn’t work, you might have to deal with another BIOS reset. 😉
Also, don’t forget that IF scales with your BCLK. For example, at 2000 IF, you’re already running around 2050 MHz. If I were you, I’d try increasing it step by step and benchmarking as you go. In my case, at 104.5 BCLK, I’m running 2067 IF, which sits at approximately 2158 MHz. I attempted pushing it higher, but unfortunately didn’t have any luck. For a day to day use, I think I'll keep it @104 bclk
I think I'm just about done tweaking the CPU on this board. I'm just wondering how far it could have gone with an Eclk board...tried 105bclk, but Y-cruncher wouldn't pass nor cinebench r23.
The 8088 is quite a relic! I think the F-14 Tomcat could have had CADAC, based on something similar. LOL! My beginnings were with the ZX-80, Commodore 64, MSX, and Apple IIc (didn't own one but was messing with it in the school workshop ), followed by the Amiga. Of course, we Amiga fanboys always knew that PCs weren't meant for gaming! LOL :D

Then it all started with the i486, followed by the Pentium 60, and so on.
You make some good points, but I'm not chasing ultra overclocking, so I think I'm fine where I am for now. I'll see how it goes with new BIOS updates. My 4090 doesn't like anything above 102.5; that's its limit. Any higher causes a black screen during boot, and it's clear it can't handle it. I'm surprised it actually holds up at 102.5 , Thanks man!
 
I'm also much more experienced with Intel CPUs, but I’ve spent an enormous amount of time these past months tweaking the 9800X3D, my fourth AMD CPU. Before that, I think I owned around 12–14 Intel CPUs. (Yes, I'm old enough for that! 😉) My first PC was hosting an Intel 8088 processor (IBM PC), running at 5 or 8 MHz if I remember correctly, and I was about 10 years old at the time.
To answer your question, a higher Infinity Fabric frequency would typically enhance both single-thread and multi-thread performance. However, keep in mind that if it doesn’t work, you might have to deal with another BIOS reset. 😉
Also, don’t forget that IF scales with your BCLK. For example, at 2000 IF, you’re already running around 2050 MHz. If I were you, I’d try increasing it step by step and benchmarking as you go. In my case, at 104.5 BCLK, I’m running 2067 IF, which sits at approximately 2158 MHz. I attempted pushing it higher, but unfortunately didn’t have any luck. For a day to day use, I think I'll keep it @104 bclk
I think I'm just about done tweaking the CPU on this board. I'm just wondering how far it could have gone with an Eclk board...tried 105bclk, but Y-cruncher wouldn't pass nor cinebench r23.
install the new chipset drivers — they slightly boost gaming performance. They just came out a few days ago.
 
CPU-Z version 2.15 did you try that one (that's the one I'm using) ? 2.13 is also known to be ok .

The A32 already came with that fix, so I feel like they're not telling us everything. It seems like they're trying to address certain issues, but they can't really endorse any changes unless they're confident those changes work. We're essentially the guinea pigs here. That said, the A33 seems pretty solid. To be honest, I haven't exactly been kind to MSI in the past, but I think they've done a great job with this one, until ... I say the opposite ;)
To add information from my previous test of the A33, it seems that, for CPPC parameters, the driver value provides the best stability and delivers great overall numbers in benchmarks.
EDIT : For my latest testing tonight, I managed to regain single-thread performance with a 104.5 BCLK(only trade off, was setting my DDR5 speed lower, it seems no matter what voltage I throw in, it would not accept more than 8360 mhz). These sessions weren’t particularly long, so I can't guarantee it’s 100% stable yet. However, it successfully passed tests using OCCT, Cinebench R23, AIDA64, CPU-Z, and Y-cruncher (which showed the most impressive improvement), even 3dmark is showing me some love in the numbers it displays ;) .
PS: disturbing how low I can set CO curve, -46 right now (-44 was maximum with A32 and -36-37 with A25)
2.11 is the one that comes bundled with the MSI Driver Installer.
It seems like 2.15 doesn’t break my system.


If that’s really the case — and I’m not the only one having issues — MSI should probably remove that version from their online installer.
 
Back
Top