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 testing was going well until this morning.. another lockup at windows boot where I had to press the reset button on the PC. On 2nd boot it came up just fine. Going to clear CMOS, reflash same A33 BIOS, clear CMOS again and leave all settings at default and see what happens. Will then try a fresh windows install if that does not work. After that I may try another motherboard.. I just want a stable PC at this point.

Any new Beta BIOS on the horizon Svet?
 
Came to report after the third resume-from-sleep, my PCIe seemed to be locked at 2.0, the mouse chopped up, but this time the entire system froze solid and had to be powered off at the switch and back on again. I turned on Spread Spectrum to see if stability will be any better, but after weeks of no problems and spread spectrum disabled, now the PCIe speed problems are happening more and more frequently with no hardware changes.
I'm getting to where @Indabox is and just want a stable machine... that manages to resume from sleep properly. I've not seen any boot issues, only resume from sleep.
 
Came to report after the third resume-from-sleep, my PCIe seemed to be locked at 2.0, the mouse chopped up, but this time the entire system froze solid and had to be powered off at the switch and back on again. I turned on Spread Spectrum to see if stability will be any better, but after weeks of no problems and spread spectrum disabled, now the PCIe speed problems are happening more and more frequently with no hardware changes.
I'm getting to where @Indabox is and just want a stable machine... that manages to resume from sleep properly. I've not seen any boot issues, only resume from sleep.
Sleep and Hibernate are features heavily not recommended everywhere as bugged and major trouble makers. I do not tried sleep/hibernate on my Carbon, but on my second machine, with Taichi Carrara and 7950x3D, sleep and Hibernate not causing any problems, which was a big surprise to me. Today updated bios to A33, no visible differences, computer working like before, flawless.
 
Came to report after the third resume-from-sleep, my PCIe seemed to be locked at 2.0, the mouse chopped up, but this time the entire system froze solid and had to be powered off at the switch and back on again. I turned on Spread Spectrum to see if stability will be any better, but after weeks of no problems and spread spectrum disabled, now the PCIe speed problems are happening more and more frequently with no hardware changes.
I'm getting to where @Indabox is and just want a stable machine... that manages to resume from sleep properly. I've not seen any boot issues, only resume from sleep.
While it is your absolute right to use sleep or hibernation, it is a part of the Windows environment and hardware forever. I don't use it as Idont trust it even more with ongoing projects, I rather save and shut down then leave it to sleep or in hibernation as in the past I lost some crucial project due to this fuckery. Now that said, as it is a feature of Windows and hardware, it should be recommended and usable, and by this day, it should work without a glitch, considering how long the feature has been available. Possibly the reason I have never run into this PCIe downgrade on my system could be because I never use sleep or hibernation.
 
The other day I was leaving for work, I needed to quickly boot up my computer to print out a contract. Normally booting up takes 15 seconds but on this day, the dreaded stutter thing happened not only on first boot but also on the second after a restart. God I was annoyed.
 
The other day I was leaving for work, I needed to quickly boot up my computer to print out a contract. Normally booting up takes 15 seconds but on this day, the dreaded stutter thing happened not only on first boot but also on the second after a restart. God I was annoyed.
And of course it has to happen when you need something urgent and quick :D what a bugger!
 
While it is your absolute right to use sleep or hibernation, it is a part of the Windows environment and hardware forever. I don't use it as Idont trust it even more with ongoing projects, I rather save and shut down then leave it to sleep or in hibernation as in the past I lost some crucial project due to this fuckery. Now that said, as it is a feature of Windows and hardware, it should be recommended and usable, and by this day, it should work without a glitch, considering how long the feature has been available. Possibly the reason I have never run into this PCIe downgrade on my system could be because I never use sleep or hibernation.
I always save before sleeping, but I don't like to wait for booting cycles so sleep the pc between sessions.
It absolutely should work. And my last system (x570) did it perfectly.
That being said, I wouldn't be surprised it's any combination of BIOS issues, Windows 24h2 fuckery, or Nvidia drivers continuing to be terrible.
 
Windows sleep and other power settings never work 100 % it works sometimes and then it don't,t any one that been on windows from win95 to win 11 know not to use it . And I don't think Microsoft will ever fix it 100%. It sucks because it is a nice feature when it works.
 
Windows sleep and other power settings never work 100 % it works sometimes and then it don't,t any one that been on windows from win95 to win 11 know not to use it . And I don't think Microsoft will ever fix it 100%. It sucks because it is a nice feature when it works.
Yes, this hypocrites bragging about how they care about the environment and talking do gooders corporate BS we have plenty this days, but this extremely useful for saving energy feature they have literally in the center of their [***CENSORED***].
 
Yes, this hypocrites bragging about how they care about the environment and talking do gooders corporate BS we have plenty this days, but this extremely useful for saving energy feature they have literally in the center of their [***CENSORED***].
I don't know about all that (don't particularly think big corps give two shits about environmentalism) but I sure do like saving about 10-20% of my electric bill to have 4 family gaming computers go into low-power mode in the 8 hours per day they aren't being used - and not having to wait for POST, sign-on, and reloading and re-positioning all of my apps and windows across 2-3 monitors (depending on family member).
 
Windows sleep and other power settings never work 100 % it works sometimes and then it don't,t any one that been on windows from win95 to win 11 know not to use it . And I don't think Microsoft will ever fix it 100%. It sucks because it is a nice feature when it works.
They've been flaky in the windows 95-windows 8 days, but actually worked quite well in Win10 if you had proper hardware and drivers installed. MS has done such a poor job of Win11 though that I wouldn't be surprised at all if they had regressed significantly.
 
Update: As of now, my system is finally stable and no longer locking up from a cold boot. GPU PCIE is also staying locked at 16x Gen5 each boot. I failed simple troubleshooting of changing 1 thing at a time to pinpoint the issue, so I'm not sure exactly what resolved my problems. My guess is that this was not a Motherboard/BIOS issue at all.. I'm leaning towards Windows/Nvidia nonsense..

Anyway, here is what I did.. either one of these or combination of all of this appears to of fixed my boot issues...

1. Opened windows event viewer system log and noticed I was receiving a lot of DCOM 10016 warnings at boot. All related to various applications being unable to launch due to a couple system accounts not having correct permissions. After a lot of google searching, I was able to correct those and now have a completely warning/error free event log at boot. This took multiple hours and was not a simple thing to do..
2. Cleared CMOS, Re-flashed BIOS A33, Cleared CMOS again, and then configured all my BIOS settings
3. Set spread spectrum from "auto" to "enabled" in BIOS
4. Updated NVIDIA drivers from 572.47 to the latest 576.28 (Used DDU for a clean install)

No issues cold booting anymore and PCIE staying locked at 16x Gen5.. (tested through multiple cold boots). In fact, when I hit the BIOS screen showing windows loading (spinning circle) it loads much quicker now. Cinebench R23 1 hour multi core test passed, Testmem5 passed, OCCT stability test passed.. and with all core PBO undervolt -30 and Expo enabled.

I have no idea why I had all those DCOM warnings. I clean installed WIN 11 24H2 Pro when building the PC and installed very few applications.. (mostly tools like CPU-Z, GPU-Z, benchmark tools etc..) so they must have been there from beginning.

Either way, things are looking good but will check back in if the system acts up again...
 
Last edited:
Update: As of now, my system is finally stable and no longer locking up from a cold boot. GPU PCIE is also staying locked at 16x Gen5 each boot. I failed simple troubleshooting of changing 1 thing at a time to pinpoint the issue, so I'm not sure exactly what resolved my problems. My guess is that this was not a Motherboard/BIOS issue at all.. I'm leaning towards Windows/Nvidia nonsense..

Anyway, here is what I did.. either one of these or combination of all of this appears to of fixed my boot issues...

1. Opened windows event viewer system log and noticed I was receiving a lot of DCOM 10016 warnings at boot. All related to various applications being unable to launch due to a couple system accounts not having correct permissions. After a lot of google searching, I was able to correct those and now have a completely warning/error free event log at boot. This took multiple hours and was not a simple thing to do..
2. Cleared CMOS, Re-flashed BIOS A33, Cleared CMOS again, and then configured all my BIOS settings
3. Set spread spectrum from "auto" to "enabled" in BIOS
4. Updated NVIDIA drivers from 572.47 to the latest 576.28 (Used DDU for a clean install)

No issues cold booting anymore and PCIE staying locked at 16x Gen5.. (tested through multiple cold boots). In fact, when I hit the BIOS screen showing windows loading (spinning circle) it loads much quicker now. Cinebench R23 1 hour multi core test passed, Testmem5 passed, OCCT stability test passed.. and with all core PBO undervolt -30 and Expo enabled.

I have no idea why I had all those DCOM warnings. I clean installed WIN 11 24H2 Pro when building the PC and installed very few applications.. (mostly tools like CPU-Z, GPU-Z, benchmark tools etc..) so they must have been there from beginning.

Either way, things are looking good but will check back in if the system acts up again...
I'm still really disturbed by A33. My performance, like I've said in many threads before, was on par with my overclocking at 104 BCLK (with A25 BIOS), but now I've moved up to 105.25 BCLK, hitting a whopping 5709 MHz on max boost,yet performance-wise, I'm not impressed.

I've passed all stability tests,OCCT, Cinebench, AIDA, etc.,without an ECLK. I didn't think it was possible to hit such a high BCLK. I do see spikes at 1.4V on CPU Vcore, but honestly, I don't really care,if it dies, it would have been a good death, and I’d have a solid reason to grab a 9950X3D. ;)
 
Last edited:
I'm also seeing more stability with the system overall. Had a PCIe speed lockup even with Spread Spectrum enabled, but with the newest 576,28 nVidia drivers as well as the upgraded RealTek sound drivers posted on the motherboard page, it seems to have helped at least.
Still have the occasional lockup in Dyson Sphere Program but I'm guessing that's the game itself or still lingering issues with the nVidia driver. We'll see how long it lasts.
 
I'm also seeing more stability with the system overall. Had a PCIe speed lockup even with Spread Spectrum enabled, but with the newest 576,28 nVidia drivers as well as the upgraded RealTek sound drivers posted on the motherboard page, it seems to have helped at least.
Still have the occasional lockup in Dyson Sphere Program but I'm guessing that's the game itself or still lingering issues with the nVidia driver. We'll see how long it lasts.
This is really strange,passes every test I throw at it, aces all benchmarks, but only the single-thread performance improves. The multi-threaded performance is actually a bit slower than at 104 BCLK, about 1% lower.
Remember when people used to say you had to burn in a CPU to get the most out of it? I always thought that was just a myth, but now…
Or maybe my CPU is putting on its final act before it dies.
1746461866186.png

And this is what HW64info says :
1746462548941.png
 
Last edited:
This is really strange,passes every test I throw at it, aces all benchmarks, but only the single-thread performance improves. The multi-threaded performance is actually a bit slower than at 104 BCLK, about 1% lower.
Remember when people used to say you had to burn in a CPU to get the most out of it? I always thought that was just a myth, but now…
Or maybe my CPU is putting on its final act before it dies.
View attachment 201623
And this is what HW64info says :
View attachment 201624

Higher overclocking seems to impact real-time performance negatively on 9800X3D, causing more game stuttering, instability, and crashes. It appears that a certain point results in diminishing returns and poor overclocking expected results. I'm quite sure 9950X3D will behave the same possible even worse due to higher core count and more temp to deal with. Anyway, good luck in "cooking" your 9800x3D, that said 5.7+Ghz is quite impressive although overall results may be disappointing at that speeds.
 
Last edited:
Higher overclocking seems to impact real-time performance negatively on 9800X3D, causing more game stuttering, instability, and crashes. It appears that a certain point results in diminishing returns and poor overclocking expected results. I'm quite sure 9950X3D will behave the same possible even worse due to higher core count and more temp to deal with. Anyway, good luck in "cooking" your 9800x3D, that said 5.7+Ghz is quite impressive although overall results may be disappointing at that speeds.
Yes, it definitely affects multi-thread performance. I’ve noticed no gains in that area starting around BCLK 102.5-103 (depends on BIOS version) and even slightly worse performance starting at 104.5, though only by about 1-3%. However, single-thread and memory performance keep improving at 107 BCLK, and I haven’t noticed any stuttering. That’s probably because, as I increase the BCLK, I also tweak down memory speed, Infinity Fabric, as well as keeping my memory from reaching impossible speeds and ensuring fabric doesn’t exceed 2200 and the CO curve.
I feel like that last setting is often overlooked in overclocking, you really need to adapt it. From my experience, I stick with the same CO between 100-103.5 (-40), then adjust to -45 for 104-106, and -50 for 106.5 and above. Mainly due to the motherboard's excessive voltage delivery when increasing Bclk on Auto or using AMD Overclocking Auto. A fixed voltage setting is rarely effective on that chip, it needs to stretch its legs and settle down from time to time.
It has been 2 days at that frequency without any crash or stutter in my benchmarks, sim racing games, or Assassin's Creed. However, the boost speed is rarely above 5688 MHz. I do get a few spikes at 5729 MHz, but 5750 MHz only occurs in benchmarks.Notice that at 107 BCLK, the real clock with max boost and PBO should be around 5800 MHz, but it never really gets anywhere near that speed.(only once I saw it boosts at 5779 mhz, but for a very brief moment)
You do have to know your memory speed and Infinity Fabric's max operating speed to achieve that. This also implies testing those settings until you encounter crashes or stuttering. From my tries, I've learned that setting the memory speed too high mostly leads to stuttering, while pushing the fabric speed too high results in crashes.

PS: Note that these high BCLK values were impossible to attain with BIOS A2xx, 105 BCLK was the last frequency at which the system would post with those BIOS versions. But with BIOS A33, I’m curious to see the max BCLK I can reach. So far, I’ve hit 107.5, but I haven’t tried going higher....yet ;)
Also you're probably right, the 9950x3D should not be as fast game wise, but multi-thread wise it's a monster.

Edit: At 107.5 it does spikes @ 5831 mhz !
1746548135262.png

But only single-thread is over the roof :
1746548219927.png


@108 bclk , everything still stable, but it wouldn't pass Y-cruncher for now. Amazing how high that board can go without destroying my nvme drive... just did a scandisk to make sure, and absolutely no error on the drive Oo.
 
Last edited:
Back
Top