MEG Z890 ACE Black Screen Crashes

V

vaerotht15ec02eb

New member
Joined
Apr 9, 2026
Messages
23
So I have had this system since Oct of 2025. It's a 285K, MEG Z890 ACE, Team Group 48Gb 8000Mhz(which is on the list of supported RAM) MSI Gaming Trio OC 5090 and 18Tb of NVMe space. After having the PC build for about a month, I would start Black Screen Crashing. No BSOD with code and errors to rely on in the Event Viewer. After I switched the RAM to 6000Mhz, the system because stable. Now I am one of those whom check the site to see if there are updates available for Driver and BIOS and have kept the BIOS updated regularly. Currently on the AA0 version. Crashes have started again. Not matter what i try, I can hardly play the PC for more than 15-20 mins.
Just incase it was the RAM, I do have a 64Gb Gskill 6000Mhz EXPO(for my previous AMD build) and it still crashes. I've ran MemTest86 on all kits and they all pass. Now after some digging my buddy had done, he saw something about eSIO updates that possibly could be applied to fix this issue. Problem is I can't find anything regarding an eSIO update for this particular board. There was another thing he brought to my attention about MSI boards(I have used ASUS for decades prior) where you have to set the Gear and then manually select the speed for your RAM and set the voltage. Tried this yesterday and was able to game all day yesterday with 0 crashes after. Now today, I am back to crashing all the time. Sometimes now after 5 mins into a game like Starfield or Star Citizen.
What other things can I try to make this system stable?
 
Last edited:
Sort by date Sort by votes
citay said:
Spinning rust = hard disk instead of SSD. Your Seagate BarraCuda 4TB. It's called spinning rust because it stores data on coated aluminium platters that spin at high RPM. If it's external, ok.



Yeah, the cooling looked excellent even in the sensors, so it can't be a cooling issue.



Ah yeah, it was only showing the primary XMP, my bad. Then all is well there.

But now you need 3.5 hours to trigger the issue, that is not good for troubleshooting. Previously it was under 20 minutes. I mean, it's an improvement on the stability side, but makes it more laborious to rule things out.

On the surface, I can't see any red flags with this system. The hardware is high-end, and sure, the high-end parts tend to be a bit more problematic sometimes (at least CPU/GPU/RAM), because they tend to be pushed to the extreme, which is fast, but not always ideal in other ways. The RAM shouldn't be an issue. The CPU and GPU seem to run at their normal specs, which especially for the GPU are quite ridiculous, but that's how it's designed. Do you use a riser card for the GPU, by the way? Actually, I would be interested in a photo of the system, it should be something to behold.

Worst comes to worst, you'd open a ticket with MSI and see what they suggest. Luckily, both board and GPU are from MSI.
Click to expand...
2.jpg
1.jpg
 
Upvote 0 Downvote
Now a buddy of mine said he had similar issues and that it ended up being the CPU. A issue with the Memory Controller Chip. And that that was giving him black screen crashes too.
 
Upvote 0 Downvote
vaerotht15ec02eb said:
Now a buddy of mine said he had similar issues and that it ended up being the CPU. A issue with the Memory Controller Chip. And that that was giving him black screen crashes too.
Click to expand...

Try updating the firmware.
kr.msi.com

MEG Z890 ACE | Gaming Motherboards｜Best Motherboard for AI PC｜MSI

The MEG Z890 ACE Intel LGA1851 gaming motherboard features Ultra Connect, including Wi-Fi 7, 5G LAN and high-speed Thunderbolt. The latest DDR5, PCIe 5.0, EZ DIY and M.2 Gen5 make it an ultimate gaming motherboard for AI PCs and future upgrades.
kr.msi.com kr.msi.com
 
Upvote 0 Downvote
Tourica said:
Try updating the firmware.
kr.msi.com

MEG Z890 ACE | Gaming Motherboards｜Best Motherboard for AI PC｜MSI

The MEG Z890 ACE Intel LGA1851 gaming motherboard features Ultra Connect, including Wi-Fi 7, 5G LAN and high-speed Thunderbolt. The latest DDR5, PCIe 5.0, EZ DIY and M.2 Gen5 make it an ultimate gaming motherboard for AI PCs and future upgrades.
kr.msi.com kr.msi.com
Click to expand...
The BIOS? BIOS is already at AA0. Which is the newest version. The only firmware that MSI has listed for my Board is a "PD FW". Which i believe was only for USB issues. That's already been applied too.
 
Upvote 0 Downvote
Now that's a nice system. Me being all-Noctua too, I can certainly appreciate going for those nice Noctua fans instead of the usual RGB-heavy affair one tends to see in such cases. Plenty of airflow, just not sure about the short distance to the wall behind, but the photos may look a bit misleading. And I guess a lot of the air is also exhausted from the top.

vaerotht15ec02eb said:
Now a buddy of mine said he had similar issues and that it ended up being the CPU. A issue with the Memory Controller Chip. And that that was giving him black screen crashes too.
Click to expand...

But wouldn't it throw RAM errors then at DDR5-6000 still? Or was his RAM stable and he still got the issues? Not so sure what can happen with Core Ultra, we don't see them terribly often on the forum.


ericj159802e9 said:
Have there been cases where platters would rust? I assumed it to be a mere metaphor. .Just thinking out loud. You'd expect HD enclosure (close) to be air tight with PCs not being wet devices. But let's not drag this topic off the rails.
Click to expand...

Let's just drag it off just a tiny bit: I didn't know this fact (or don't remember it), but early HDDs actually used iron oxide as the coating, see here.
So that would be spinning rust in the quite literal way!
 
Upvote 0 Downvote
citay said:
Now that's a nice system. Me being all-Noctua too, I can certainly appreciate going for those nice Noctua fans instead of the usual RGB-heavy affair one tends to see in such cases. Plenty of airflow, just not sure about the short distance to the wall behind, but the photos may look a bit misleading. And I guess a lot of the air is also exhausted from the top.
Click to expand...
Yeah, it's about 6" from the wall. I know I should pull it away a little more just so heat just doesn't sit and cycle back into the case. Kind of an unfortunate size to my office

citay said:
But wouldn't it throw RAM errors then at DDR5-6000 still? Or was his RAM stable and he still got the issues? Not so sure what can happen with Core Ultra, we don't see them terribly often on he forum.
Click to expand...
My last crash was when it was at 6000Mhz. This has been the odd thing about this situation. I can go a week or two with absolutely no issues, at all. Then, once I get a black screen crash, it'll happen every time I launch a game. Some games it takes like 5-10 mins into a game starting and sometimes it crashes as the game is starting from desktop. Every once in a while, the crashes will happen when just starting Brave or Chrome.

The only thing I can think of is once it started getting warmer in my area and of course my system running a little bit hotter, that's when i started seeing these crashes happen more consistently. Now, during my last "heat" wave it was about 84f-91f(29c-33c), this was the last time this really got bad and just crashes all the time and I just couldn't do anything. System never really got hotter than 75f(24c). This is the only correlation I can think of. And since I want to start streaming and recording stuff, I can't have the system crash when it gets a little warm. Otherwise I can never do much. lol Now for the last 5-6 days it's been on average 55f(12c) so heat hasn't been a concerned. But in my area during summer it gets to 115f(46c) sometimes.
 
Upvote 0 Downvote
The worst thing here is the long interval between crashes and that you can't provoke it reliably. This complicates things.

Normally I would say, try to rule out the GPU next. Either by borrowing a different one (difficult to adequately replace that monster you got there somehow, I know), or hell, you could even try taking out the GPU entirely and using the iGPU via the HDMI out of the board. But what games can you play with an Intel iGPU... something like Civilization, or very undemanding games. But if you say, it starts even when you just launch a game, then why not, try it with that iGPU.

Because if you can completely rule out the GPU, then the focus would go more towards the CPU indeed.
 
Upvote 0 Downvote
citay said:
The worst thing here is the long interval between crashes and that you can't provoke it reliably. This complicates things.

Normally I would say, try to rule out the GPU next. Either by borrowing a different one (difficult to adequately replace that monster you got there somehow, I know), or hell, you could even try taking out the GPU entirely and using the iGPU via the HDMI out of the board. But what games can you play with an Intel iGPU... something like Civilization, or very undemanding games. But if you say, it starts even when you just launch a game, then why not, try it with that iGPU.

Because if you can completely rule out the GPU, then the focus would go more towards the CPU indeed.
Click to expand...
To rule out my 5090, I did try a few weeks back with my 4090. Had crashes then too. I wanted to rule out my most expensive thing first a long time ago. lol
 
Upvote 0 Downvote
Hmm ok. But with the focus now more on the CPU, it's difficult to test. What you can check, in the BIOS there should be a setting somewhere called "Intel 200S Boost" or so, make sure that's on [Disabled].
 
Upvote 0 Downvote
Well, I think I figured out what is creating the crashing for months now...6 month old Gaming Trio OC 5090. At least i still have my 4090 and haven't sent that to my friend yet.
 

Attachments

  • ugh.jpg
    ugh.jpg
    1.8 MB · Views: 33
Upvote 0 Downvote
Oh wow. Yeah that's not normal at all. But didn't you say you had crashes with the 4090 too?

What happens if you run FurMark 2, does it also show artefacts of some kind?
 
Upvote 0 Downvote
I realized it crashed with the 4090 when I was at 8000Mhz.

citay said:
Oh wow. Yeah that's not normal at all. But didn't you say you had crashes with the 4090 too?

What happens if you run FurMark 2, does it also show artefacts of some kind?
Click to expand...
I tried running different tests when this first happened. Just starting most programs crashes the system before it fully loads. The image prior was after CP2077 was booting up and after hitting the spacebar the first time, it locked up before I could even hit Space the second time so i couldn't even hit the main menu.
 
Upvote 0 Downvote
Now, here is a question. If you look at the image where the window is off of my case, you can see I am using a CableMod 90 degree connector cable. This is not an extension, it's a complete cable. This is their 600w rated cable so it's designed to work with the 5090. I removed that and put the SeaSonic cable back up but I also put it on the LianLi Vertical GPU Mount because there is no way to have the factory style connector attach to any GPU purely for the fact that I was thinking the cable may be the issue. That's when i took it all apart again, went back to directly to the board, still had this pixelation and crashing.
Strix.jpg
 
Upvote 0 Downvote
vaerotht15ec02eb said:
I realized it crashed with the 4090 when I was at 8000Mhz.
Click to expand...

Ah ok, yes, makes sense if the memory system wasn't stable at that speed.

vaerotht15ec02eb said:
Now, here is a question. If you look at the image where the window is off of my case, you can see I am using a CableMod 90 degree connector cable. This is not an extension, it's a complete cable. This is their 600w rated cable so it's designed to work with the 5090. I removed that and put the SeaSonic cable back up but I also put it on the LianLi Vertical GPU Mount because there is no way to have the factory style connector attach to any GPU purely for the fact that I was thinking the cable may be the issue. That's when i took it all apart again, went back to directly to the board, still had this pixelation and crashing.
Click to expand...

The power cable cannot cause these image artefacts. A riser cable or vertical GPU mount could cause it, because they worsen the PCIe signal quality, it's just a question of how much they worsen it. But if you tested it in both orientations and with two cables, then it should come more from the card itself.

You can check in the BIOS, set PCI_E1 to Gen4 speed, see if that does anything. PCIe 4.0 is easier to achieve in terms of PCIe signal quality than 5.0, obviously. So if it's in any way related to that, things might improve. But I don't know, sounds more like the card itself, maybe the VRAM.

These artefacts, can they appear in any game? Or just one particular one? Because if it's only in one game, then it might be a 50-series-related bug in there.
 
Upvote 0 Downvote
I tried launching many games multiple times and it all led to crashing. This was the only time I could get into a game menu to see this and then it crashed. I'm going to reach out to MSI directly and see what they say. They may just want to have the card sent in to RMA. Not sure though. I've been gaming on this 4090 all morning so far and not one issue. Not even a stutter. So I am going to continue to play and if by later on, all that gaming stress hasn't made the system crash, then we know(or easier to speculate) it's the GPU creating the system issue. I may try and install the 5090 again and see if setting the PCIe speed will help, but at this point it's a little more obvious what the issue is. And the LianLi riser cables are typically pretty good. Not thin cheap cables. I used this with my 4090 a while back and the performance was always the same compared to being directly attached to the motherboard.
 
Upvote 0 Downvote
So this was after at least 5 hours of gaming and 1 hour of OCCT while watching YouTube.
1776020822907.png

Big difference is that when i would do OCCT and have YouTube open, just to apply more stress, having the 5090 installed the system, it would crash. I have not had one crash happen since putting this 4090 back in my system.
 
Upvote 0 Downvote
That narrows it down a lot to the 5090, yes. BTW, you can disable the Intel iGPU in the BIOS if you don't use it. Usually under Settings - Advanced - Integrated Graphics, set "Initiate Graphic Adapter" to "PEG" and set "IGD Multi-Monitor" to [Disabled], then it's off.
 
Upvote 0 Downvote
Back
Top