ASRock Rack B650D4U MicroATX for Ryzen 7000 processors

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

suchar

New Member
Sep 1, 2024
2
0
1
I have the same problem: board bought in June and bricked itself last week. First it rebooted itself several times, then it stopped post-in at all (dr debug code 00). BIOS and BMC were not upgraded recently (BIOS is 10.15, BMC is 4.10.0), IPMI is still working but shows hardware as powered off, and power on command does not work (fans and hard drives start spinning but nothing else happens). Tried reseating memory and cpu to no avail. When booting with a single memory stick, the IPMI interface still shows two memory sticks, so it looks like hardware is not read at all.

Contacted asrock support, looks like a lot of theses board are failing recently :(
 
N

newborn2010

New Member
Jul 16, 2018
13
1
3
suchar said:
I have the same problem: board bought in June and bricked itself last week. First it rebooted itself several times, then it stopped post-in at all (dr debug code 00). BIOS and BMC were not upgraded recently (BIOS is 10.15, BMC is 4.10.0), IPMI is still working but shows hardware as powered off, and power on command does not work (fans and hard drives start spinning but nothing else happens). Tried reseating memory and cpu to no avail. When booting with a single memory stick, the IPMI interface still shows two memory sticks, so it looks like hardware is not read at all.

Contacted asrock support, looks like a lot of theses board are failing recently :(
Click to expand...
I've tried flashing the firmwares and clearing the CMOS, but neither worked. From what I've observed, most of the reports about systems getting stuck at 00 have occurred in the past few months, so I'm wondering if a relatively new firmware could have caused damage to a component on the motherboard? For instance, with ASUS's early Dominus motherboards, there was a widespread issue where systems would get stuck at 00 during boot, and the only way to start the system was by heating it. Eventually, someone discovered that the problem was due to an unreasonable resistor value design on the motherboard, making it extremely sensitive to temperature and unable to boot. Replacing the resistors resolved the issue. Could it be that the B650 motherboard is facing a similar issue, where the newer BIOS firmwares might be causing excessive voltage/current, leading to the damage to a component?
 
K

krdx

New Member
Aug 31, 2021
3
0
1
I have two B650D4U boards that I bought at roughly the same time.

Same config: Ryzen 7700 with 32GB ECC ram (2x 16GB Kingston 9965800-015.A00G ). Not using the built-in NICs besides IPMI. Temps for cpu are around 50C, ram 33C with good airflow - 1U chassis. Cpu has a copper cooler with a fan. Dual PSU setup. Running debian with 6.x kernels.

M80-GC025700xxx v4.0.1 (bios 3.11, bmc 3.10.00, power-on hours 138 days) - So far stable

M80-G9007100xxx v4.0.1 (bios 3.11, bmc 3.10.00, power-on hours 150 days) - Stopped working 4 days ago. I Logged remotely via ipmi, there was no video output/no post, I could power on/off it but nothing was happening, so I left it turned off.
Went and physically removed the server, turned it on, it worked. Hammered it with memtest for a day, but no errors with various ram configs. So I thought maybe it was an accident/ram slot was not properly seated.
Plugged it back again and after two days (today) it rebooted itself twice already for no reason. Then I came across this topic.

Should I bother with updating BIOS/BMC in hopes that this may resolve this issue, or as other posters said, most likely common hardware problem and I should RMA the board/look for a replacement?
 
Last edited:
N

newborn2010

New Member
Jul 16, 2018
13
1
3
RolloZ170 said:
have you tested the CPU with another board ?

possible CPU burned.
Click to expand...
That's what I firstly thought too. I applied for an RMA for the 7950X3D. After installing the brand new 7950X3D that came back from the RMA, it still got stuck at code 00."
 
O

orangejuiced

New Member
Sep 2, 2024
1
0
1
I've had 3 B650D4U boards die in the past 3 weeks. All 3 were bought in/around June. Starts off as some random crashes/reboots. Couple days after the boards start showing code 00 and no longer boot. IPMI still works but power options fail. Tried swapping out the memory, CPU and power supply to no avail. Replacing the board is the only fix for now.

Will have to check what firmware and bmc version the dead boards are running. Very concerned more are going to start dying in the coming weeks.
 
D

dmz

New Member
Sep 6, 2024
1
0
1
We've had the same issues with tons (30+) of of the 2L2T/BCM versions, no issues on regular B650D4U however. Where did you guys source the B650D4Us with issues?
 
N

netswitch

Member
Sep 24, 2018
70
36
18
Mostly at a Dutch Dealer, Azerty (www.azerty.nl).
How did you manage the RMA of your faulty boards ? Directly with AsrockRack ?
Here we did an RMA with Azerty but they could not see the issue.. (which I can understand as it sometimes takes 10 days to crash).

All our faulty boards are "revision 4.01" (printed on the board next to the cpu socket)
 
K

krdx

New Member
Aug 31, 2021
3
0
1
krdx said:
M80-G9007100xxx v4.0.1 (bios 3.11, bmc 3.10.00, power-on hours 150 days) - Stopped working 4 days ago. I Logged remotely via ipmi, there was no video output/no post, I could power on/off it but nothing was happening, so I left it turned off.
Went and physically removed the server, turned it on, it worked. Hammered it with memtest for a day, but no errors with various ram configs. So I thought maybe it was an accident/ram slot was not properly seated.
Plugged it back again and after two days (today) it rebooted itself twice already for no reason. Then I came across this topic.
Click to expand...
Just wanted to say that this mobo died today with dr. debug code 00, but this time it doesn't post at all no matter what :(
 
J

jei

Active Member
Aug 8, 2021
201
128
43
Finland
B650D4U REV 4.01 running since april. So far no problems but debating getting a new board as I would not prefer much downtime for this box.

BMC 04.10.00
BIOS 10.15
 

Attachments

N

newborn2010

New Member
Jul 16, 2018
13
1
3
I sent the motherboard(B650D4U-2L2T/BCM REV3.03) I brought second-hand to a third-party repairer and they fixed it right away today, costing about 80 dollarso_O. When I asked the agent which module was faulty, the response was vague, only saying it was an issue with the BGA circuit. Since I could not get the details of the malfunction, I have to send another B650D4U-2L2T/BCM REV4.01 to the local distributor who will ship it to the manufacturer in Taiwan. It may take up to two months.
 
Last edited:
F

FrankTL

Member
Aug 9, 2022
78
30
18
newborn2010 said:
I sent the motherboard(B650D4U-2L2T/BCM REV3.03) I brought second-hand to a third-party repairer and they fixed it right away today, costing about 80 dollarso_O. When I asked the agent which module was faulty, the response was vague, only saying it was an issue with the BGA circuit. Since I could not get the details of the malfunction, I have to send another B650D4U-2L2T/BCM REV4.01 to the local distributor who will ship it to the manufacturer in Taiwan. It may take up to two months.
Click to expand...
Sounds like they tried some generic fixes like reballing the bga chip(s) and this fixed the issue
 
R

RolloZ170

Well-Known Member
Apr 24, 2016
10,718
3,435
113
germany
FrankTL said:
Sounds like they tried some generic fixes like reballing the bga chip(s) and this fixed the issue
Click to expand...
i don't think so. you will see the removal of the BGA and resoldering them clearly.
preheating will mostly destroy the caps or reduce the lifetime at least.
you will never get a socket (BGA) replacement for only $80usd thought, reballing etc will cost must more.
 
B

bpye

New Member
Apr 13, 2021
15
2
3
Does anyone know if enabling SRV-IO for the on-board BCM NIC is possible on the B650D4U-2L2T/BCM? I've enabled SR-IOV support, as well as all the PCIe ARI options I can find but no dice. sriov_totalvfs in Linux shows the configured number of VFs, but when I attempt to add a VF with sriov_numvfs I'm getting the typical "type 7f class 0xffffff", "unknown header type 7f, ignoring device" errors which seemingly indicate that ARI is disabled.

My hunch is that this is because the BCM NIC is downstream of the B650 chipset, but maybe I've missed something...

If that's the case I guess I'll swap to the B650D4U without the integrated BCM NIC, that way I can install an SR-IOV capable NIC in the CPU x4 slot, and I get the chipset x4 M.2 slot to use for additional storage instead.
 
Last edited:
J

Joze Volf

New Member
Jun 6, 2018
3
0
1
Another one here, B650D4U rev 4.01. Server failed in less than 2 months, I ordered new board and replaced it, put it back in to production and has now failed again in <2 months. I received a new board as warranty replacement for the first one but now I do not want to replace it again and put it into production because I expect it to fail again in <2 months. CPU is Ryzen 9 7900, nothing extreme in TDP.
 
You must log in or register to reply here.
Top Bottom