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.

Chriggel

Active Member
Mar 30, 2024
147
80
28
Why would you need to install Windows 10 to virtualize Server 2003? The host OS doesn't have to do anything with the guest OS.

For your other questions, hmm, difficult to say. These questions make even less sense than the first one. Maybe a little language barrier problem here? Or a crucial misunderstanding.

All I can say at the moment is that you probably shouldn't run Server 2003 under any circumstances. It's old, it's unsupported and if you still need it for whatever reason, you have a really big problem.
 

Chriggel

Active Member
Mar 30, 2024
147
80
28
You take the virtualization platform of your choice, create a virtual machine and install Server 2003 from an installation disk image.
 
Jan 3, 2023
64
29
18
The question is why would you want to, unless it is also for running deprecated software that requires Windows Server 2003.
 
Jan 3, 2023
64
29
18
I wanted to comment on my ongoing experience with my X13SAE and the 13900K, especially given the reports of instability of these processors with gaming motherboards.

I run my X13SAE with PL1 and PL2 settings of 230W and 350W respectively. The original PL1 I started with was 200W, and after monitoring temps and other parameters carefully, I raised PL1 to 225W and then to 230W over a year ago.

My machine does distributed computing tasks for scientific computing, including BOINC and folding@home, which keeps both the CPU and 4080 GPU 100% busy almost constantly.

I have experienced ZERO crashes or instability with this setup for over 1.5 years of operation.

My guess is that Supermicro is conservative and has all of the proper current excursion protections, etc. enabled, unlike the gaming motherboards, so the CPU is super stable. Note I use the machine for gaming occasionally, and have experienced ZERO crashes there too.

Perhaps part of it is that I am running OpenSUSE Leap 15.5, which is also super stable.
 

saf1

Member
Nov 27, 2022
56
10
8
I wanted to comment on my ongoing experience with my X13SAE and the 13900K, especially given the reports of instability of these processors with gaming motherboards.

I run my X13SAE with PL1 and PL2 settings of 230W and 350W respectively. The original PL1 I started with was 200W, and after monitoring temps and other parameters carefully, I raised PL1 to 225W and then to 230W over a year ago.

My machine does distributed computing tasks for scientific computing, including BOINC and folding@home, which keeps both the CPU and 4080 GPU 100% busy almost constantly.

I have experienced ZERO crashes or instability with this setup for over 1.5 years of operation.

My guess is that Supermicro is conservative and has all of the proper current excursion protections, etc. enabled, unlike the gaming motherboards, so the CPU is super stable. Note I use the machine for gaming occasionally, and have experienced ZERO crashes there too.

Perhaps part of it is that I am running OpenSUSE Leap 15.5, which is also super stable.
I have a similar experience with no failures although I am not pushing mine as you are. Mine is used for storage and Plex. I used to see some heavy CPU utilization when a family member had to transcode but that went away once I added in the Nvidia Quadro P400. Much more efficient transcoding leaving the CPU idle.

Every once in a while I will fire up some Windows VM's using Ubuntu's LXD implementation but other than that the X13SAE and 13900K is overkill for my use case. However it has worked excellent once I sorted through the HD connection issues I ran into.
 
Jan 3, 2023
64
29
18
Some commentary on the Raptor Lake (Intel 13th and 14th gen) CPU's and the Supermicro X13SAE(-F) motherboards.

Supermicro had, for a few days, BIOS version 3.3, which included the 0x125 microcode update, which fixed eTVB (enhanced Thermal Velocity Boost) in the i9 parts. The website has now, temporarily reverted back to 3.1. My guess is that they are testing the just released 0x129 microcode update, which caps the Vcore supply to 1.55 V to prevent the processor ring-bus damage that was occurring because of Vcore voltages that were too high during single or dual thread operation of the CPU (i.e. where the clock speed of the P-Cores would be the highest and need the most voltage).

Until then, if you are worried about voltages above 1.55 V being provided to the CPU before (or even after the BIOS with 0x129 comes out), you can install the following two files in vcore_log_service.zip that provide a script based monitoring service:

The vcore_log.service goes into the /etc/systemd/system directory. The actual script, record_vcore.sh, goes into the /usr/local/bin directory.

You can enable the service after putting the files in place by systemctl daemon-reload then systemctl enable vcore_log.service and systemctl start vcore_log.service.

The script uses the bc console calculator, and the rdmsr tool from msr-tools. you may have to install packages for those commands if they are not available on your system.

The voltage maximum is set by editing record_vcore.sh and changing the vcore_log value. It is currently set to the Intel recommended maximum for Raptor Lake of 1.55 V. Note that after a BIOS upgrade is done with the 0x129 microcode version, you should see NO entries.

The script will log an entry to /var/log/vcore.log and also to the dmesg if this voltage is exceeded. It samples every second.

Note that I have been running this script for several days on my X13SAE with a 13900K and OpenSUSE Leap 15.6 and have NO entries. I have BIOS 3.3 with the 0x125 microcode installed.
 

Attachments

  • Like
Reactions: chenxiaolong

Weapon

Active Member
Oct 19, 2013
317
116
43
Some commentary on the Raptor Lake (Intel 13th and 14th gen) CPU's and the Supermicro X13SAE(-F) motherboards.

Supermicro had, for a few days, BIOS version 3.3, which included the 0x125 microcode update, which fixed eTVB (enhanced Thermal Velocity Boost) in the i9 parts. The website has now, temporarily reverted back to 3.1. My guess is that they are testing the just released 0x129 microcode update, which caps the Vcore supply to 1.55 V to prevent the processor ring-bus damage that was occurring because of Vcore voltages that were too high during single or dual thread operation of the CPU (i.e. where the clock speed of the P-Cores would be the highest and need the most voltage).

Until then, if you are worried about voltages above 1.55 V being provided to the CPU before (or even after the BIOS with 0x129 comes out), you can install the following two files in vcore_log_service.zip that provide a script based monitoring service:

The vcore_log.service goes into the /etc/systemd/system directory. The actual script, record_vcore.sh, goes into the /usr/local/bin directory.

You can enable the service after putting the files in place by systemctl daemon-reload then systemctl enable vcore_log.service and systemctl start vcore_log.service.

The script uses the bc console calculator, and the rdmsr tool from msr-tools. you may have to install packages for those commands if they are not available on your system.

The voltage maximum is set by editing record_vcore.sh and changing the vcore_log value. It is currently set to the Intel recommended maximum for Raptor Lake of 1.55 V. Note that after a BIOS upgrade is done with the 0x129 microcode version, you should see NO entries.

The script will log an entry to /var/log/vcore.log and also to the dmesg if this voltage is exceeded. It samples every second.

Note that I have been running this script for several days on my X13SAE with a 13900K and OpenSUSE Leap 15.6 and have NO entries. I have BIOS 3.3 with the 0x125 microcode installed.
Thanks for the post. I went to update my BIOS and noticed the version on Supermicro’s website was lower than what I have installed.
 

BlueMatt

New Member
Apr 10, 2024
2
0
1
Looks like they re-uploaded the 3.3 BIOS as 3.3a but it still only includes the 0x125 microcode update.
 
Jan 3, 2023
64
29
18
Supermicro tends to be super conservative on BIOS updates. My guess is that they are running pretty intensive internal validation on the 0x129 microcode.
 

BlueMatt

New Member
Apr 10, 2024
2
0
1
I assume so, but also 0x129 presumably is important to reduce processor aging, so even in the worst case if it doesn't reduce instability at all it seems like something worth shipping without being 100% sure it removes all instability.
 

unwind-protect

Active Member
Mar 7, 2016
540
216
43
Boston
I assume so, but also 0x129 presumably is important to reduce processor aging, so even in the worst case if it doesn't reduce instability at all it seems like something worth shipping without being 100% sure it removes all instability.
Yeah, but they will be sneaking in other changes to the BIOS, and that needs testing.
 
Jan 3, 2023
64
29
18
3.3b seems to work fine. No detectable performance changes. Note that I was not seeing any issues with Vcore above 1.55v according to my logging script/service I posted above even before 3.3b with the 0x129 microcode.
 

rekd0514

Member
May 11, 2015
40
14
8
Does anyone have the 3.1 BIOS file for the X13SAE-F? My unraid server has been crashing every few days since I updated to 3.3 and as far as I know that is all I have changed. I would like to test that theory by reverting back. I didn't see any location to download these short of contacting Supermicro.

I read Intel is releasing a new microcode again anyway (0x12B). Intel Rolls Out Third Major "0x12B" Microcode Patch To Fix 14th & 13th Gen CPU Instability Issues

"During its ongoing investigation, Intel discovered another issue with Vmin shift in affected CPUs, which can cause the motherboard and BIOS code to request elevated voltages during idle or light activity periods."

This part makes me think the instability in the low power modes is causing the crashing because I am using powertop in unraid to reduce idle power consumption on the 13600k.
 

chenxiaolong

New Member
Nov 16, 2020
10
6
3
Does anyone have the 3.1 BIOS file for the X13SAE-F? My unraid server has been crashing every few days since I updated to 3.3 and as far as I know that is all I have changed. I would like to test that theory by reverting back. I didn't see any location to download these short of contacting Supermicro.

I read Intel is releasing a new microcode again anyway (0x12B). Intel Rolls Out Third Major "0x12B" Microcode Patch To Fix 14th & 13th Gen CPU Instability Issues

"During its ongoing investigation, Intel discovered another issue with Vmin shift in affected CPUs, which can cause the motherboard and BIOS code to request elevated voltages during idle or light activity periods."

This part makes me think the instability in the low power modes is causing the crashing because I am using powertop in unraid to reduce idle power consumption on the 13600k.
The Internet Archive didn't archive the files themselves (Wayback Machine), but if you open the browser's dev console and search for `fwdownloadtarget` on the archived pages, you can grab the ID and pass that to Supermicro's actual site to download the older files.

* BIOS_X13SAEF-1C48_20230406_2.1_STDsp.zip: https://www.supermicro.com/support/resources/getfile.php?SoftwareItemID=17685
* BIOS_X13SAEF-1C48_20231005_3.0_STDsp.zip: https://www.supermicro.com/support/resources/getfile.php?SoftwareItemID=19688
* BIOS_X13SAEF-1C48_20231225_3.1_STDsp.zip: https://www.supermicro.com/support/resources/getfile.php?SoftwareItemID=20216
* BIOS_X13SAEF-1C48_20240826_3.3b_STDsp.zip: https://www.supermicro.com/support/resources/getfile.php?SoftwareItemID=22453
 

chenxiaolong

New Member
Nov 16, 2020
10
6
3
My X13SAE-F unfortunately just died yesterday (of course, not long after the warranty expired...). The IPMI still works perfectly, but judging from the PSU relay clicking sounds, the main system is resetting every second. I did try another known-good PSU and saw the same behavior with that, so I'm guessing the motherboard is probably dead.

I'm not sure what caused it, but I did start to see weird issues after upgrading the BIOS to 3.3a initially and then 3.3b. When saving BIOS settings, it was occasionally complaining about lack of NVRAM space. Certain things just silently failed to save, like custom secure boot keys. I didn't see any large EFI variables in `/sys/firmware/efi/efivars` at the time so perhaps something else was eating up NVRAM space.
 

custom90gt

Active Member
Nov 17, 2016
249
99
28
40
My X13SAE-F unfortunately just died yesterday (of course, not long after the warranty expired...). The IPMI still works perfectly, but judging from the PSU relay clicking sounds, the main system is resetting every second. I did try another known-good PSU and saw the same behavior with that, so I'm guessing the motherboard is probably dead.

I'm not sure what caused it, but I did start to see weird issues after upgrading the BIOS to 3.3a initially and then 3.3b. When saving BIOS settings, it was occasionally complaining about lack of NVRAM space. Certain things just silently failed to save, like custom secure boot keys. I didn't see any large EFI variables in `/sys/firmware/efi/efivars` at the time so perhaps something else was eating up NVRAM space.
Odd about the NVRAM space issue. Tried clearing the CMOS? Other steps like reseating the ram?