Pro Z790-A WiFi + DDR5 6400 XMP Stability Issues with Recent A4 Bios

[WoodburyMan]

Greetings!

I have the Pro Z790-A WiFi paired with a i9 13900K.

I am also having stability issues with DDR5 memory. I have been trying to run DDR5 6400 memory in XMP mode with varying degrees of stability. (2 x 32gb sticks kit).
The memory I have specifically is G.Skill F5-6400J3239G32GX2-TZ5RK ( https://www.gskill.com/product/165/374/1665644504/F5-6400J3239G32GX2-TZ5RK which is DDR5-6400 CL32-39-39-102 1.40V ). The memory is listed on MSI's compatibility site as compatible, and on G.Skill's site. I'm running it with a Cooler Master MasterAir MA612 Stealth with a decent 6-Pipe heatsink. A nice even thermal paste spread of Thermal Grizzly Kryonaut was put on. On benchmarks maxes out at 85C, games and real use scenario like 75C. I don't run any overclocking on the CPU, just default BIOS settings, aside from XMP mode. I have also tried witching my new PSU for this build out with a known good PSU from my last build with the same results.

Running A3 or earlier BIOS, the system was somewhat stable running in XMP mode with 6400MT stock XMP timings and voltage. I would get occasional hiccups with notoriously glitchy titles like COD. Some random odd behavior but usable stable. However after the A4 BIOS update, it's downright unable. I noticed in the changelog there were memory compatibility changes. I once in a great while get a BSOD in Windows, related to system stability, some more narrow as Clock Watchdog Timeout, but mostly in games get random exit errors. (COD:MWII and Fortnite being biggest offenders). On A4 BIOS the only stable setting is running stock 4800MT settings. I have not tried any other settings besides that and XMP, and if anyone had any pointers who may have this issue, any timings that worked for you would be helpful.

Wondering what others results with DDR5 6400 is in XMP mode. I see others have success running it straight up, others having issues, perhaps due to the processor bin lottery with onboard controller variations. Any specific timing settings, clockrates, BIOS settings to try that are known stable for others with XMP 6400 2x32gb memory would be VERY helpful.

Also, I have run memory86+, memtest86, and Windows Memory diagnostic on it in 4800MT stock settings with zero errors. Running on 6400MT XMP profile, I get random single series errors here and there which lead me to believe the memory is fine, just an issue of timings.

TIA

 

[oktay.fati158102e0]

What's your exact G.Skill RAM kit model name? When there is a boot delay, which one of the "EZ Debug LEDs" is on (they are above the 24-pin ATX header)? Also, which LED is on when the PC doesn't want to boot at all? What is your exact SSD model, CPU cooler model, PSU model?

"What's your exact G.Skill RAM kit model name?"
G.Skill Ripjaws S5 DDR5-5600 32gbx2
İntel XMP Ready / CL36-36-36-89 /1,25 V

"When there is a boot delay, which one of the "EZ Debug LEDs" is on (they are above the 24-pin ATX header)? Also, which LED is on when the PC doesn't want to boot at all?"
At startup, it blinks white, then orange. Then none of them light up.
Suddenly after 30-35 seconds there is a beep sound and Logo screen.

"What is your exact SSD model,"
1-MLD M300 NVMeSS 500gb (system installed)
2-Samsung SSD 840 EVO Replacement
3-Toshiba SSD TR200 Replacement

"CPU cooler model, "
Deep Cool Gammaxx Series LE500 240 mm Liquid CPU Cooler

"PSU model?"
REcom Power Engine 750CF (750 W)

Thanks again

 

[oktay.fati158102e0]

What's your exact G.Skill RAM kit model name? When there is a boot delay, which one of the "EZ Debug LEDs" is on (they are above the 24-pin ATX header)? Also, which LED is on when the PC doesn't want to boot at all? What is your exact SSD model, CPU cooler model, PSU model?

First of all thanks for the answer. I wrote a reply about 2 hours ago, but now when I refresh the page I don't see it. This reply may appear 2 times. I apologize.
RAM
G.Skill Ripjaws S5 DDR5-5600 32gbx2
İntel XMP Ready / CL36-36-36-89 / 1,25 V
EZ Debug LEDs
In both cases, White and orange blink at first and then none of them light up. A little while ago the system received an update and wanted to restart the computer a few times but all attempts failed and in the meantime I observed that no led did not make any sign.
SSD Model
1-MLD M300 NVMeSS 500gb (system installed)
2-Samsung SSD 840 EVO Replacement
3-Toshiba SSD TR200 Replacement
CPU cooler model
Deep Cool Gammaxx Series LE500 240 mm Liquid CPU Cooler
PSU model
Recom Power Engine 750CF (750W)
Thanks again.

 

[citay]

irst of all thanks for the answer. I wrote a reply about 2 hours ago, but now when I refresh the page I don't see it.

The forum probably put it onto page 2 automatically and you were still on page 1. It's ok.

REcom Power Engine 750CF (750 W)

This immediately sticks out. A cheap no-name PSU with a Z790 / 13600 KF / DDR5 system? It doesn't match at all! These kinds of PSUs are usually quite bad. It's not about the wattage, this is probably enough (what's your GPU, forgot to ask?), but the quality is too low. It doesn't even have any 80PLUS certificate whatsoever, this is no good.

I would replace this with a higher-quality PSU right away, see my Guide: How to find a good PSU. It doesn't matter if it fixes the problem or not. Because that no-name PSU is not a good foundation for your new system. But if it fixes it, you kill two birds with one stone (i hope the translation works for this idiom).

MLD M300

This is also a bit weird, a no-name SSD, and it's the boot drive. But since your problems happen before Windows starts to boot, it's probably not involved.

 

[oktay.fati158102e0]

Thanks @citay
Yes, my power supply is old and I couldn't budget for it.
I read your article about RAMs with a link in your signature. I benefited a lot. It was a very good explanatory article. After a 4-hour test with the program you mentioned there called "memtest86", the RAMs were successful. This was a consolation for me.

 

[citay]

If the RAM is error-free with XMP enabled, that's a good sign, so it might not be involved. But your PSU, especially if it's old, you cannot trust it at all. I looked it up, it has two years of warranty, which for a PSU is laughably short, it basically means, the company has no confidence in the longevity of this PSU. On top of that, it uses an outdated internal design and has too low efficiency for today's standards. This is not good at all for a system like yours. This part i would immediately replace with a better one. If you link me a shop or two where you normally order your hardware, i can take a look at what PSU models they offer that would be good for your system.

The last part you want to save money on is the PSU. I would save everywhere else before i would save on the PSU. If you get a slower CPU, nothing much will happen, just the performance will be a bit lower. Same for the RAM and GPU. But if you save on the PSU, it stresses all the attached components, it can even damage them in case of a failure, because important protection circuits are often missing on these cheap PSUs. And you can have all kinds of weird problems that you cannot explain. This PSU you have is like a ticking timebomb. And maybe your problems even come from there.

 

[oktay.fati158102e0]

Hi @citay
I'll buy it in the next few days. I actually ordered this, but I had a problem with the company and postponed it.
"High Power Performance GD 800W 80+ Gold (HP1-J800GD-F12S)"
I don't know if it's right or not to link it. So I just wrote the name of the PSU. I also read your article about choosing a PSU. I'll make a choice in the light of all this information. The markets here are not stable. There is a wide variety of markets. If there is a model you recommend, I will look for it here.
Thank you very much.

 

[citay]

High Power Performance GD 800W is a Turkey-only PSU model, but it seems to be pretty decent, it uses good technology. Also, "High Power" is not no-name (despite the weird name), they are an OEM supplier for many other brands, and also sell PSUs under their own brand, similar to Seasonic. So this would already be a decent PSU model to get. Could be better, but could also be much worse. Your 3080 Ti can draw 350W of power, and the 13600KF can be thirsty too, so yeah, a 750-800W PSU is not really overkill here.

Assuming you are indeed in Turkey, this is one PSU i would recommend:

Be Quiet! Pure Power 12M 750W 80+ Gold Tam Modüler Güç Kaynağı-BN343 Fiyat

Be Quiet! Pure Power 12M 750W 80+ Gold Tam Modüler Güç Kaynağı-BN343 uygun fiyat ve hızlı teslimat Gelbura.com

www.gelbura.com

https://www.hepsiburada.com/be-quiet-pure-power-12m-750w-80-gold-tam-moduler-guc-kaynagi-psu-bn343-pm-HBC00004MH5YE

It's probably more expensive than the High Power, but it's also better. It has 10 years manufacturer's warranty, this is truly a good model.
But even the "High Power" would be a much-needed improvement from your current PSU.

 

[oktay.fati158102e0]

Hi @citay That's way out of my budget. I've had worse problems. I bought the other one and the problem persists. And I don't know what to do anymore.

 

[citay]

You said in another thread that your SSD is no longer detected? Edit: Ah, that was another user. Anyway, take out all drives (yes, i know it can't boot Windows anymore then). This is to see if any of the drives are causing the long boot delay before Windows even starts to boot. If the delay is gone, add the drives back one by one to see which one causes it.

 

[oktay.fati158102e0]

Yes, before in the boot menu the name of the SSDs appeared but now it doesn't and I can't recognize the system disk and I can't enter the system. It keeps going to BIOS. Now the boot menu shows the name of the system disk but not the other disks.

 

[oktay.fati158102e0]

It's been about 10 days and I haven't gone crazy. My problems continue.
I want to write my problem again. Because I'm going to make a decision. I'll either return the motherboard. but it's going to be hard. So I'll just throw it in the bin and get rid of it.

First of all, can I find out, is this the official forum of MSI? Are the problems posted here read by MSI? Or is it just help between members?

Because there are similar threads. Problems are not solved in those threads either. What is MSI doing about these problems? Do they take any action? Should I open a new thread to draw attention? Don't get me wrong, I just can't find my way out.

I've given up on the other late startup problems. My biggest problem is; when installing and updating the system, the system restarts in "reboot" situations and after a few seconds of operation, the system freezes. I manually reboot and the same thing happens. After a few attempts, the system reboots with the message "not properly installed changes are being undone", the system reboots again, freezes again, reboots again, freezes again..... I am in such a cycle. I have removed all unnecessary devices. I have used NvMe M2, SSD and classic 3.5 and 2.5 inch Hard Discs. I changed my power supply. I changed the video card. I changed the NvMe M2 disc. I updated the BIOS (A90) and the system became completely unstable. I'm back to A70.

I can't determine which hardware is causing this, there is no hardware left anyway. I've tried all the settings in the bios, nothing. Today I was able to finish the installation after maybe 40-50 shutdowns and reboots. Now it gives a system update and reboot warning. Looks like I'll be struggling until the morning.
Thanks in advance for your attention. I have my hardware information in the signature section. I will go to another brand's motherboard tomorrow in line with your opinions.
Regards

 

[citay]

First of all, can I find out, is this the official forum of MSI? Are the problems posted here read by MSI? Or is it just help between members?

Just experienced users helping users. If you want to bring it up with MSI directly, open a ticket here, https://account.msi.com/services/tickets

Don't throw the board away (never put such devices into the bin; if anything, you would give it to electronics recycling). Open a ticket with MSI and let's see what they say.
You have the Z790-P Wifi, perhaps not the best choice to pair with a 13600K(F), from MSI i would've recommend the Z790-A Wifi instead. But this will not be the reason. Well, it can be your specific board, but it's not like the Z790-P is unstable with all 13600Ks. If you want to get a new board from MSI, get a Z790-A Wifi. Alternatives in that price range are ASUS Prime Z790-P WIFI or GIGABYTE Z790 Gaming X AX.

Today I was able to finish the installation after maybe 40-50 shutdowns and reboots.

Forget it, it's not worth using. Files will already be corrupted with such a grave instability. The system first needs to be made stable somehow, then you can install Windows.

 

[oktay.fati158102e0]

Thank you @citay
What I mean by throwing it away is not to use it, but to utilise it in some other way.
I've written directly to MSI and we'll see what they say.
Thanks again.

 

[JohnFromMGMT]

Thank you @citay
What I mean by throwing it away is not to use it, but to utilise it in some other way.
I've written directly to MSI and we'll see what they say.
Thanks again.

Have you tried just one stick of RAM in slot2?it sounds to me like you have a memory stability issue.

if it were a PSU issue your problems wouldn’t be experienced during startup unless your PSU is completely failing; it would be experienced when pushing your graphics card & CPU to max.

my fingers are crossed you’ve solved this problem by now.

 

[JohnFromMGMT]

Thank you @citay
What I mean by throwing it away is not to use it, but to utilise it in some other way.
I've written directly to MSI and we'll see what they say.
Thanks again.

I will note that my ram also passed all memtest86 stability tests but would fail when starting. Very similar symptoms to what you’re experiencing, just not so exasperated.

 

[oktay.fati158102e0]

Have you tried just one stick of RAM in slot2?it sounds to me like you have a memory stability issue.

if it were a PSU issue your problems wouldn’t be experienced during startup unless your PSU is completely failing; it would be experienced when pushing your graphics card & CPU to max.

my fingers are crossed you’ve solved this problem by now.

Thank you. I think I tried one by one without knowing it, but now that you tell me, I will try again more consciously. Unfortunately, the problem persists. The reason for my anger is that I cannot find the source of the problem. In my country, there are no companies or services to find a solution, and the return process for unpackaged products is long and complicated. I will try and write the result. Thanks again. I hope you will solve your problem too.

 

[miguely15f802ed]

Greetings!

I have the Pro Z790-A WiFi paired with a i9 13900K.

I am also having stability issues with DDR5 memory. I have been trying to run DDR5 6400 memory in XMP mode with varying degrees of stability. (2 x 32gb sticks kit).
The memory I have specifically is G.Skill F5-6400J3239G32GX2-TZ5RK ( https://www.gskill.com/product/165/374/1665644504/F5-6400J3239G32GX2-TZ5RK which is DDR5-6400 CL32-39-39-102 1.40V ). The memory is listed on MSI's compatibility site as compatible, and on G.Skill's site. I'm running it with a Cooler Master MasterAir MA612 Stealth with a decent 6-Pipe heatsink. A nice even thermal paste spread of Thermal Grizzly Kryonaut was put on. On benchmarks maxes out at 85C, games and real use scenario like 75C. I don't run any overclocking on the CPU, just default BIOS settings, aside from XMP mode. I have also tried witching my new PSU for this build out with a known good PSU from my last build with the same results.

Running A3 or earlier BIOS, the system was somewhat stable running in XMP mode with 6400MT stock XMP timings and voltage. I would get occasional hiccups with notoriously glitchy titles like COD. Some random odd behavior but usable stable. However after the A4 BIOS update, it's downright unable. I noticed in the changelog there were memory compatibility changes. I once in a great while get a BSOD in Windows, related to system stability, some more narrow as Clock Watchdog Timeout, but mostly in games get random exit errors. (COD:MWII and Fortnite being biggest offenders). On A4 BIOS the only stable setting is running stock 4800MT settings. I have not tried any other settings besides that and XMP, and if anyone had any pointers who may have this issue, any timings that worked for you would be helpful.

Wondering what others results with DDR5 6400 is in XMP mode. I see others have success running it straight up, others having issues, perhaps due to the processor bin lottery with onboard controller variations. Any specific timing settings, clockrates, BIOS settings to try that are known stable for others with XMP 6400 2x32gb memory would be VERY helpful.

Also, I have run memory86+, memtest86, and Windows Memory diagnostic on it in 4800MT stock settings with zero errors. Running on 6400MT XMP profile, I get random single series errors here and there which lead me to believe the memory is fine, just an issue of timings.

TIA

Join bios, use profile 1 XMP but go to dram voltage and try 1.45v, i think this should fix the instability that have your system.

 

[JohnFromMGMT]

Thank you. I think I tried one by one without knowing it, but now that you tell me, I will try again more consciously. Unfortunately, the problem persists. The reason for my anger is that I cannot find the source of the problem. In my country, there are no companies or services to find a solution, and the return process for unpackaged products is long and complicated. I will try and write the result. Thanks again. I hope you will solve your problem too.

Thank you for your good wishes. I have solved my instability issue and then some!

It seems like you’re thinking of the little things already but to be sure; When you try one stick at a time do it on default JDEC settings.
To eliminate any kind of PSU issues I’d go factory settings on CPU.

Eject & disconnect any backup storage you have before you try for safety- though I’d be scrupulous of any backups you have while your system is unstable anyways.

 

[oktay.fati158102e0]

Hello, guys. One of the rams was faulty. It's fine now.

Thanks again.

 

[citay]

Glad it's solved, and thanks for the feedback.

That's surprising that one RAM module was bad, because it passed Memtest86 testing. How did you find out that it was faulty?