LGA 1700 Alder Lake "Servers"

Notice: Page may contain affiliate links for which we may earn a small commission through services like Amazon Affiliates or Skimlinks.

bambi123

Member
Oct 12, 2022
31
3
8
This is going to be a bit long but I suspect the X13SAE(-F) has some sort of stability issue, and I think it is related to the BIOS.

Installed Server 2019, and everything worked great - even Hyper-V which is not meant to work with the E-cores enabled. Installed all drivers and moved our VM's over and it has been working fine for around 7 days.

1st problem, on a reboot to add features to the host OS, some of the VM's would not start - the VM configuration files some how got corrupted. We re-created the VM's and attached the original VHDX. We did many disk checks / scans nothing wrong with the NVME drives that store the VM's (they are not new moved from the previous server).

2nd problem, the OS install got corrupted, and Server 2019 started to moan that sources could not be found to install a role - tried all the fixes, even adding the sources manually, etc, etc, So to save time - I installed a new boot SSD, and installed a fresh copy of Server 2019, exactly as before - enabled Hyper-V and rebooted, now getting a BSOD - weird, so I disabled the E-cores to see if that would work, and yes it did - so Hyper-V does need the E-cores disabled to work, so why did it work before with the E-cores enabled? (This is why I am thinking some sort of BIOS issue).

3rd problem, the fresh install of 2019 went well, everything installed, imported the VM's checked they work ok - then started the windows updates to bring it up-to-date, Then got a BSOD on reboot, with stopcode 0000021a, - tried everything to fix. So called it a day and re-installed the original BOOT SSD, and got back to where I had started the day.

Like I said the system works great when running, reboots scare me now, - just cannot trust if it will boot up with no errors.

Action plan for next week - install another Boot SSD, update firmwares BMC and BIOS, reset the board (BMC/BIOS to default), re-apply all the BIOS settings, and re-install Server 2019 - and see if we can get it to install and reboot with no BSOD's or file corruption.

Just adding my experience.
 

Weapon

Active Member
Oct 19, 2013
289
104
43
Yeah I agree, board seems to have some sort of stability issue. Pretty disappointing after waiting this long.

I swapped back to my Ubuntu install and while it seems to be stable, I did notice the display driver seems to have disappeared or something since swapping CPUs from the 12900k to 13900k. Resolution is stuck at 1024*768 and I don't know how to force a driver update/install in linux. This is just a test install, so whatever, I can redo it.

Power supply is questionable, as it's an old 430 watt Corsair, the only spare I have on hand right now. I've never had an issue with it and the system has no dGPU so it should be enough even under max load. Also I believe the default BIOS settings are enforcing some sort of lower turbo power limit as running stress-ng on all 32 threads in linux only draws 165 watts from the wall.

This also leads me to believe the issue is not the power supply, but I have a 1000 watt on the way to test. I also swapped back in my non-ECC DDR5 and got the same BSOD in Windows. It occurs without even entering the desktop if I just sit at the login screen for a few minutes.

And one more issue, the board will not stay shut down properly. When I select shut down, it just reboots. I've looked for various BIOS settings thinking maybe "Wake on LAN" or some equivalent was on. I can't find anything and it happens even without an Ethernet cable plugged in.

I guess my next steps are to try the new PSU just in case. Would also like to try another dGPU and see if I can get that working. And finally brand new, fresh install of Windows with manual installation of latest drivers direct from Intel.

Right now feels like a long ways away from switching to this board/platform from my trusty C246 and 2288G on my main workstation.
 

Kiririn

New Member
Apr 7, 2022
17
11
3
Most of these issues sound like they stem from the integrated GPU. 21A BSOD is basically the same as the EF BSOD mentioned a few posts ago. Random file corruption could be explained by GPU driver hosing memory, but would run prime95 to rule out a more typical hardware issue first. GPU not initialising in Linux is weird assuming kernel is up to date, but supports conclusion of GPU issue. Shutdown issue sounds unrelated but nevertheless a problem
 
Last edited:

rootshell

Member
Jan 10, 2021
72
31
18
Latest news is pretty disappointing from someone who was about to order. Hopefully you guys figure out the root cause, as now I'm hesitent to jump in.
 
  • Like
Reactions: Kiririn

Weapon

Active Member
Oct 19, 2013
289
104
43
Any suggestions on what BIOS settings to change to get full performance out of my 13900K? I’m only seeing like 27K in R23
 

bambi123

Member
Oct 12, 2022
31
3
8
Most of these issues sound like they stem from the integrated GPU. 21A BSOD is basically the same as the EF BSOD mentioned a few posts ago. Random file corruption could be explained by GPU driver hosing memory, but would run prime95 to rule out a more typical hardware issue first. GPU not initialising in Linux is weird assuming kernel is up to date, but supports conclusion of GPU issue. Shutdown issue sounds unrelated but nevertheless a problem
I am not using the built GPU of the CPU, so is it worth me disabling the GPU?

But the question then becomes in my case could also be the ASPEED 2600 GPU / DRIVER causing the problem rather than the CPU GPU.
 

yeryer

New Member
Sep 24, 2022
23
4
3
I've been running a X13SAE-F and 12900k under high workloads for weeks and it's been perfectly stable. It's completely headless though so if there's a GPU issue I wouldn't have hit it.

The last parts arrive tomorrow for my next X13SAE-F build with a 13900k and I'll post the benchmark results when that's complete. In the past, I have seen poor performance from an LGA 1700 chip when the CPU cooler was socket 1151. It's a longshot but something to double check. That cooler will fit and appear to mount fine but doesn't contact tight enough to allow the cpu to draw full wattage
 

Weapon

Active Member
Oct 19, 2013
289
104
43
Did you change any BIOS settings to get the full performance from your 12900k? I'm still seeing only around 2000 on Cinebench R23 single core vs 2200 for most reviews.

Also I believe I have solved the reboot vs shutdown issue by changing the BIOS setting for "power button behavior" from "resume last state" to "power off"

Still having various GPU related issues. For example. I currently have a dGPU installed and system seems stable. I powered down, switched HDMI cable from dGPU to iGPU. Power up, no video, get 5 error code beeps. Pretty inexcusable in my opinion. This is basic functionality that doesn't work. Platform has been out for like a year at this point, simply adding ECC support shouldn't result in these issues or or a $200 premium but I'm getting off topic now.
 

yeryer

New Member
Sep 24, 2022
23
4
3
I don't think I changed any BIOS settings on my12900k build but I'm not sure today. I do remember getting those 5 beeps and no video after a dozen reboots and feeling like motherboard was completely defective.
 

Weapon

Active Member
Oct 19, 2013
289
104
43
Can you check and see what your current settings are for CPU power limits? And how did you solve the 5 beep issue?
 

yeryer

New Member
Sep 24, 2022
23
4
3
I'm looking through the IPMI displays and don't think it shows that... unfortunately I can't reboot it to get into the bios. I worked around the 5 beeps by never plugging in an hdmi cable and found the IPMI address by listing devices on my router.
 

bambi123

Member
Oct 12, 2022
31
3
8
Did you change any BIOS settings to get the full performance from your 12900k? I'm still seeing only around 2000 on Cinebench R23 single core vs 2200 for most reviews.

Also I believe I have solved the reboot vs shutdown issue by changing the BIOS setting for "power button behavior" from "resume last state" to "power off"

Still having various GPU related issues. For example. I currently have a dGPU installed and system seems stable. I powered down, switched HDMI cable from dGPU to iGPU. Power up, no video, get 5 error code beeps. Pretty inexcusable in my opinion. This is basic functionality that doesn't work. Platform has been out for like a year at this point, simply adding ECC support shouldn't result in these issues or or a $200 premium but I'm getting off topic now.
I have the 12700K installed, do not have a dGPU installed, but can run CPUZ bench if you want to compare.

My PL1 and PL2 settings are defaulted on my board.
 

Weapon

Active Member
Oct 19, 2013
289
104
43
I have the 12700K installed, do not have a dGPU installed, but can run CPUZ bench if you want to compare.

My PL1 and PL2 settings are defaulted on my board.
Yes, please do. And Cinebench R23 if it's possible. Thank you!
 

bambi123

Member
Oct 12, 2022
31
3
8
Yes, please do. And Cinebench R23 if it's possible. Thank you!
Sorry cannot run cinebench the server is in use with hyper-v's running.

The CPUZ returns - 12700K single thread 770, multi thread 7514 , this is with the E-cores disabled, and 10 hyper-v VM's running.
 
  • Like
Reactions: Weapon

mrmh

New Member
Oct 3, 2022
20
1
3
X13SAE problems (BIOS v1.0b):

1. PCH temperature readings are from 100 to 108 degrees of Celsius according to BIOS
1a. it's weird because I can rest my finger on side of heatsink.
1b. HWINFO does not show PCH temperature!??
1c. I have Noctua fan with 1000 rpm running on PCH heatsink (DIY)


2. cannot upgrade BIOS because "Launch EFI shell from filesystem device" ends up with "Not Found"
2a. cannot add new boot option because only SATA disk with Linux is detected for EFI shell filesystem
2b. I can boot Windows 11 which has it's own EFI partition with EFI shells located in it
 

Weapon

Active Member
Oct 19, 2013
289
104
43
Weird. Updating the BIOS was one of the few things that worked flawlessly for me with this board. Will check to see what my PCH temps are.
 

bambi123

Member
Oct 12, 2022
31
3
8
X13SAE problems (BIOS v1.0b):

1. PCH temperature readings are from 100 to 108 degrees of Celsius according to BIOS
1a. it's weird because I can rest my finger on side of heatsink.
1b. HWINFO does not show PCH temperature!??
1c. I have Noctua fan with 1000 rpm running on PCH heatsink (DIY)


2. cannot upgrade BIOS because "Launch EFI shell from filesystem device" ends up with "Not Found"
2a. cannot add new boot option because only SATA disk with Linux is detected for EFI shell filesystem
2b. I can boot Windows 11 which has it's own EFI partition with EFI shells located in it
1. I have a Noctua cooler and fan installed, and it spins too slow, so it shows as critical on BMC Webgui.

2. If you hit F12 while booting it should launch the builtin EFI shell. You have may have removed it from the boot menu in the BIOS. But I still suspect there is somesort of BIOS bugs with this motherboard, so maybe remove the CMOS batt, and then default the BIOS - and re-do your settings.
 

bambi123

Member
Oct 12, 2022
31
3
8
I was remembering the name of the setting wrong, here is a screenshot of the actual setting that seems to have solved this issue for me
I have this set to POWER ON, in case of a power cut, the motherboard powers on when power is restored.