THX i will try it on Weekend again.
I let you know if it works
For anyone updating the BMC via Socflash:
If you have a graphics card installed, after resetting your BIOS to default settings be sure to go re-enable the D-Sub video output, and set it to the default video device a few options down. I had to do this else the socflash tool would print ā[Error] Canāt find the device.ā
Seems that perhaps the video output to ipmi is needed to get the bmc to be recognized by the flash util?
Note youāll have to plug a monitor into the dsub/vga output to continue with the flashing, or do it headless via IPMI.
2 Likes
I tried verifying I could boot to FreeDOS first before updating either the BIOS or BMC, and I got the typical message indicating that no bootable device exists selecting either the UEFI, USB, or āCDROMā options in the motherboardās boot menu. Do I have to upgrade the BIOS to 1.40 first before upgrading the BMC or being able to boot from FreeDOS at all?
You should not need to update anything to boot to freedos. You dont have to do that to update the bios either. That you can do from the bios menu or f6 at boot or from ipmi.
When you use the boot menu (f11? forget) your usb stick may show up more than once, as a uefi device and as a usb device, at least mine did. One would boot the other did not. Booting to your stick is required to flash the BMC of course. I flashed the BIOS first, then the BMC.
Hello! I hope everyone is doing well. Iāve just bought this motherboard to build a virtualization server; the first computer Iāve tried to build in 25 years.
Iāve had a great experience with the board itself, but the IO Shield has been a problem. When I tried to fit the ports through their holes, the incredibly thin sheet metal bent, and two of the little prongs broke off. Several of the prongs arenāt actually touching the ports theyāre supposed to ground.
So, questions:
-
Do I actually need the I/O shield if I get dummy plugs for all the ports Iām not using? That should eliminate the risk of short circuits on the pins, right?
-
I would like an IO shield to keep debris out. Does anyone know of a way to get a plastic replacement 3D printed? I canāt figure out how to order a replacement from Asrock Rack, in any case.
Thanks! Iām kind of stuck at this point. I have what I need to build the server, but now Iām not sure what I should do about the IO shield. Iām so very tempted just to remove it and go on without it.
Iām paranoid about more of the little prongs, which were surely weakened by my attempts to move them into position weakened even more of them. Several of them ended up so bent theyāre not actually touching anything, so theyāre not grounding anything.
I would appreciate any advice. This is not where I expected to have problems.
1 Like
The motherboard should only have one ground plane for all those connectors, so as long as at least one touches, it should be okay. Heck, it might be fine even if none touch. Just make sure itās not threatening to either short something out or injure you when you touch it.
If you look on Amazon, thereās some product called IOMesh thatāll allows cutting your own IO shield out of a plastic grid.
1 Like
Thanks!
Iāll definitely be doing a custom plastic mesh if I ever have to mess with the IO shield again. Replacing the shield now would mean removing the motherboard again, and Iād rather do several other things than that. Perhaps wrestle a bearāand not one of the cute tie-wearing ones.
If I donāt run into problems when Iāve plugged in all the rear IO connectors, I think I should be fine. I still might end up picking up some dummy plugs to seal off the unused connectors.
I got an X570D4U running today with a 5800x. It gave me POST code d0 which stands for DXE_CPU_ERROR according to the manual. So I took out the CPU to check for bent pins (they were all fine), and found I had added way too much thermal paste. Copious amounts had squeezed out all 4 sides and when I took the CPU out some paste got on the motherboard PCB, the socket, and the CPU pins (
). I spent the next two hours meticulously cleaning that up, then about 2 minutes trying one RAM stick at a time and different RAM slots. Different RAM slots turned out to be the solution which allowed the system to POST. I had put my two RAM sticks into white slots A2 and B2 when they needed to be in blue slots A1 and B1 according to the manual, which I had not consulted because it hadnāt been included in the box.
1 Like
Entering the āx570d4u-2l2tā-ownerās club soon (ordered).
What single tip or advice/wisdom you got, working with that board for some time, for anyone running that board for the first time?
1 Like
Long time watcher, but recent viewer of the board though, have a few bits to add to the discussion on this board.
Thus far, not āthatā impressed considering its cost and it all being a bit knife and fork to setup/update. Fact the build isnāt not going into colab, we likely should have used a more consumer board but they are not setup for rack mount. Anywayā¦ see what its like in a month+
Reading over the posts yesterday while setting our system up, we have our board paired with a 5950x, dynatron A24, 128gig of KSM32ED8/32ME ram (running at 3200 in bios) 2x Samsung 980 evo 500g in raid 0 all in a SuperMicro 825TQC-R740LPB 2u chassis. Running WS2019, 1.4 bios and 1.2 BMC. idles at 4.49gig.
Running R20: 10154. max core temp 67c (amb 21.8c) the dynatron can cope with burst workloads but towards the end of r20 you can see the core speed reduce. Still learning this platform so not sure if this is a direct link to temps or the cpu clocking down due to workload.
Interested to see what temps others have in artificial workloads and also there fan curve settings. Currently left ours stock. Still feel we should have ran this on water rather than air, time yet 
Runs at 60c in bios according to bios sensors. (was a discussion about this above)
Still learning both this board and the 5950x so anything else to improve the build would be a great help.
Thanks for Bogdan_Dumitru for this posts about updating BMC
EDIT: Still can not get remote bios to work via ARR server manager. Since 1.20 now get some info text at the bottom left of the page explaining types of action, but still nothing useful 
Now gives an error: āThe resource at the URI /redfish/v1/Systems/Self/Bios was not found.ā
Thank youā¦ took me a long time, and I wouldnāt have figured out how to flash the BMC with socflash without this. I should also note (at least for me) that setting the D-Sub to primary got rid of an otherwise annoying boot time beeping
For other 5950x users with this board, is anyone else having instability issues? Iāve had to turn off Core Performance boost in bios and lock it at 3.4ghz. if I turn core performance boost back on my system crashes after several hours of use, regardless of temps/workloads. Iām on the latest drivers/bios.
No problems with that here as yet.
What kind of work loads? Our build still isnt in production so if its a public accessible workload, i am more than happy to run for x time and see if its a workload or hardware issue
Do you run your memory with supported frequencies? You can see the table in the manual
Yeah bought 2x 32gb kingston ecc that was listed in the manual, running at its regular 3200mhz, tried swapping out the sticks and the problem persists. as for workloads its mostly just a simple nas running plex, next cloud, valheim server, but it will crash at idle with none of the services running as well. If others havenāt had this issue with CPB, then Iām kinda leaning towards it being a faulty cpu
I initially thought the new version was okay, but now it seems to refuse to give USB keyboard control (buffer is perpetually full, at least once the OS boots). Hereās what syslog coming from the BMC showsā¦
EDIT: Looks like the BMC REEEEEEALLY doesnāt like USB autosuspend. Now itās working after I disabled USB autosuspend for the virtual keyboard+mouse.
Still wish I could make the virtual USB ethernet either become usable or just go away!
Just plugged in my board. The Chipset temperatures are out of control. Chassis is still open, I just wanted to check if everything is connected and detected.
x570 chipset sensor goes >90Ā°C and after log sent me my first messages (temperature too high! event) in my fancy new IPMI board, I pulled the plug for now. No load, just accessing IPMI.
I get airflow going once I install my case fans, but do you guys have the same problems with chipset temps? normal? manageable with airflow?
edit: Fixed. With some airflow after full assembly, Iām down to ~55Ā°C
Iām surprised it would get to 90C with an open chassis. I guess the fact this board doesnāt have a dedicated chipset fan like almost all other x570 boards means it actually relies on case airflow for chipset cooling. I just checked and mine is a steady 56C in a RSV-L4500U chassis, all 15 HDDs installed, stock fans. Machine otherwise idle.
I was monitoring temps and 90Ā°C was not the limitā¦linear increase until I powered down at 96Ā°C. All well with some airflow. That chipset is more demanding and hot than my CPU
Got 2x120mm fans at 1100rpm, but airflow isnāt directed at the board. As long as there is movement in the case, the chipset seems happy (for now). Installing Proxmox right now. Low load ~58Ā°C now here.
Got some trouble with my HDDs not being detected by the BIOS. Need a beer or two. They have power, but BIOS doesnāt like them. User error, all fine!
Iāve just got mine up and running, 5950x with 128GB ECC. Iāve noticed some instability issues in Windows, but nothing as severe as what youāre talking about (games only). I havenāt used Windows in about a decade, however, and my memory is that Windows was always unstable. Once I get Linux back up and on it, Iāll know more how I feel about instability