help with gigabyte MZ73-lm0 motherboard and genoa-x 2x 9684x help
- Thread starter Venturi
- Start date
Notice: Page may contain affiliate links for which we may earn a small commission through services like Amazon Affiliates or Skimlinks.
What hardware configuration did you have it set to that completed a successful POST? Go back to that and wait till it POST and attempts to boot.
Debugging primarily is conducted on minimal configurations - removing any possible failure points to minimize potential impact.
Therefore normally it would be single CPU with one-two RAMs in right slot (as per manual) without any PCIe/USB/M.2/SSD etc. devices, sometimes swapping either CPU (if you have two) or RAM in case one of those components are defective.
And yes - as others have mentioned : falling back to last known working configuration is another good option (and then starting changing components trying to figure out at what point it starts failing).
I'm currently having the same issue:
CPU: AMD EPYC Genoa-X 9384X 32-Core 3.1GHz Processor (only 1 CPU)
MotherBoard: Gigabyte MZ73-LM0 rev 2.0 AMD EPYC 9004/90 motherboard
PSU: CORSAIR RM850x Fully Modular Low-Noise ATX Power Supply – ATX 3.1 Compliant – PCIe 5.1 Support – Cybenetics Gold Efficiency – Native 12V-2x6 Connector
Ram: A-Tech 64GB DDR5 4800MHz PC5-38400 ECC RDIMM 2Rx4 (EC8 10x4) Dual Rank 1.1V ECC Registered DIMM 288-Pin
Add-On: None at this time
CPU Cooler: SilverStone Technology XE04-SP5 Black Fan Shroud 4U Server/Workstation CPU Cooler for Socket SP5, SST-XE04-SP5B
Current attempts:
Currently, I have 1 CPU installed with a CPU cooler. I have 1 64GB stick of RAM installed in slot A0. The power supply is connected only to CPU 0 and the motherboard 24-pin connector. I also have 5 case fans daisy-chained and connected to the motherboard. No PCIe cards or add-ons and no NVMe/HDD installed.
I think it is the RAM from the repeated bios codes, but I am unsure. If anyone might know, I'd greatly appreciate any solutions or things I can try to get the machine past the "Please wait for the chipset to initialize" screen.
CPU: AMD EPYC Genoa-X 9384X 32-Core 3.1GHz Processor (only 1 CPU)
MotherBoard: Gigabyte MZ73-LM0 rev 2.0 AMD EPYC 9004/90 motherboard
PSU: CORSAIR RM850x Fully Modular Low-Noise ATX Power Supply – ATX 3.1 Compliant – PCIe 5.1 Support – Cybenetics Gold Efficiency – Native 12V-2x6 Connector
Ram: A-Tech 64GB DDR5 4800MHz PC5-38400 ECC RDIMM 2Rx4 (EC8 10x4) Dual Rank 1.1V ECC Registered DIMM 288-Pin
Add-On: None at this time
CPU Cooler: SilverStone Technology XE04-SP5 Black Fan Shroud 4U Server/Workstation CPU Cooler for Socket SP5, SST-XE04-SP5B
Current attempts:
- I tried unplugging all the PCIe Add-Ons.
- I tried reseating the RAM modules.
- Reseated the CPU
- I tried updating Bios multiple times from BMC (Before and after updating BMC)
- Unplugged all HDD, SATA connections, and the NVME drive
- Tried different combinations of RAM modules (1, 2, 4, and 6)
- tried removing all power cables that are related to processor 1 only leaving processor 1 power cables connected.
- I attempted to use the BIOS recovery mode feature using the switch on the motherboard. ( I could not get this to work or understand what was needed to make it work)
- Tried clearing CMOS
- Let the computer run for 7 hours.
- Only get a screen that says, "Please wait for chipset initialize..." (this happens immediately. White static appears at the bottom of the screen, and then the text is shown)
- There is a flashing light where the m.2 is located.
- For the P0/P1 power LED, the 7 light is pulsing while lights 1, 3, and 4 are constantly on (this happens immediately when turning on the computer)
- I'm getting a repeated bios code cycle of AD, CC, DE, and 00 (not sure if this is the correct order)
- BMC works, kind of. The update bios feature does not work, but the update BMC feature does. Also, changes to bios are not persistent from BMC and revert just after closing the window or turning off and on the computer. Fix: the bios does update after dumping bios and comparing bytes. It is currently on the R04_F32 bios.
Currently, I have 1 CPU installed with a CPU cooler. I have 1 64GB stick of RAM installed in slot A0. The power supply is connected only to CPU 0 and the motherboard 24-pin connector. I also have 5 case fans daisy-chained and connected to the motherboard. No PCIe cards or add-ons and no NVMe/HDD installed.
I think it is the RAM from the repeated bios codes, but I am unsure. If anyone might know, I'd greatly appreciate any solutions or things I can try to get the machine past the "Please wait for the chipset to initialize" screen.
"DE AD xx xx" ??? ( yes looks like Dead and means exactly that )
you can download the POST codes in BMC WebGUI.
you will see it after successfull POST. there are some gigabyte with dual BIOS, so the (universal)BMC FW can not know which is active.
Last edited:
You say you left the computer running for 7 hours so I'm going to assume it was on the chipset initializing screen during this time.
If not, then you need to leave the computer running on that screen for a long while. Like up to half an hour.
If it doesn't go past that screen then it's probably the RAM.
Also there is a gigabyte mz73-lm0 rev 2.x. I have one as well. It appears the mz73-lm1 has 1.x and 3.x with no 2.x version.
If not, then you need to leave the computer running on that screen for a long while. Like up to half an hour.
If it doesn't go past that screen then it's probably the RAM.
Also there is a gigabyte mz73-lm0 rev 2.x. I have one as well. It appears the mz73-lm1 has 1.x and 3.x with no 2.x version.
Yes, the computer was running for 7 hours during the initializing screen (I should have made that more clear). I have placed an order for QVL RAM, and hopefully, that will resolve the issue (this is my current plan). I could be wrong; however, I don't think this board has a dual BIOS (this is something I'll look into to be safe).
Does the DE AD code indicate that the BIOS is dead or that the BIOS died after the CC code? (these are the last codes before repeating:
..., 0x32, 0x3B, 0x6D, 0xED, 0x00, 0x4F, 0x60, 0x61, 0xB7, 0xB6, 0x9A, 0x62, 0xDE, 0xAD, 0x00, 0x00, 0xCC, 0xCC, 0xDE, 0xAD)
Does the DE AD code indicate that the BIOS is dead or that the BIOS died after the CC code? (these are the last codes before repeating:
..., 0x32, 0x3B, 0x6D, 0xED, 0x00, 0x4F, 0x60, 0x61, 0xB7, 0xB6, 0x9A, 0x62, 0xDE, 0xAD, 0x00, 0x00, 0xCC, 0xCC, 0xDE, 0xAD)
Last edited:
the BMC Firmware is universal for all Gigabyte motherboards with AST2600 chip, so if not yours, some other can have dual BIOS.
the BIOS inventory is updated by the BIOS/CPU after successful POST
no POST = old data form last working config.
"0x00, 0x00, 0xCC, 0xCC"
is the AGESA error code capsulated in "0xDE 0xAD"
there will be no further action after this.
Update: I got it to post, but only if 1 CPU is installed and the CMOS clear pins are bridged (I'm not sure why this works, but not bridging the clear CMOS pins leads to a DE AD message). However, 2 CPUs don't work, and it gets stuck on BIOS code 28 every time. I've tested both CPUs in slot zero, and they both work independently, just not together. Has anyone experienced this or knows of a possible solution to this problem?
Hi everyone,
I’ve got my system up and running, but I’m seeing an odd issue with core clock speeds on my EPYC 9115 CPU. While the max clock speed for this CPU is 4.1GHz, only half of the cores are reaching that speed. The other half remain stuck around 600MHz, even under stress.
Initially, all cores were locked at 600MHz. However, after enabling the following BIOS setting, now only half the cores exhibit this throttling:
You can clearly see the split:
cpu MHz : 602.886
cpu MHz : 602.874
...
cpu MHz : 4120.603
cpu MHz : 4120.623
Does anyone know what might be causing this? Is this a BIOS/firmware quirk, or is there another power management setting I might be missing?
Thanks in advance for any insights!
I’ve got my system up and running, but I’m seeing an odd issue with core clock speeds on my EPYC 9115 CPU. While the max clock speed for this CPU is 4.1GHz, only half of the cores are reaching that speed. The other half remain stuck around 600MHz, even under stress.
Initially, all cores were locked at 600MHz. However, after enabling the following BIOS setting, now only half the cores exhibit this throttling:
Here’s what I’ve tried so far:
- Toggled SMT (Hyperthreading) on and off
- Changed DIMM configurations (1, 2, 4, 6, and 12 RDIMMs per CPU)
- Tested with only one CPU installed
- Increased PPT and TDP limits significantly
- Set all performance-related BIOS settings to "High Performance"
- Checked CCD Control (currently set to Auto, and seems unchangeable)
- BIOS version update
watch -n 0.5 "cat /proc/cpuinfo | grep 'MHz'"You can clearly see the split:
cpu MHz : 602.886
cpu MHz : 602.874
...
cpu MHz : 4120.603
cpu MHz : 4120.623
Does anyone know what might be causing this? Is this a BIOS/firmware quirk, or is there another power management setting I might be missing?
Thanks in advance for any insights!
Hope I can use this thread for my case: MZ73-LM2, 2x 9755 ES, 2x Samsung 64GB DDR5-6400 sticks (M321R8GA0EB2-CCPKF), Power LC1650 PSU, latest BIOS, latest BMC. When running with 1 CPU I see only "Please wait for chipset to initialize...". Few hours went by, still the same. I assume this is because either:
1. the ES CPU doesn't support 6400 speeds or
2. because M321R8GA0EB2-CCPKF (which I'm using) is not listed, but e.g. M321R8GA0EB2-CCPKC is (last 3 letters are different) in the QVL. A strange thing is in BMC under DIMM inventory it sees 96GB installed (but I use 1x 64GB stick) and 6x 16GB (DIMM_P0_A0 till DIMM_P0_F0). ChatGPT says that sometimes this can happen even with most recent BMC and BIOS versions, so I don't think much of it but plan to test with a DDR5-5600 stick.
A whole different problem is running 2 CPUs: if I turn on the PSU, the fans spin briefly and then the PSU makes a sound, like the over current protection would kick in. Now I don't know how much power everything sucks at the boot, but would expect a 1650W PSU (Power LC1650 ) could handle it (also tested on a 1700W PSU (Enermax 1700) with same result). So I wonder what kind of PSUs are people using if normal ATX PSUs get triggered by their over current protection circuits (if that's the case).
One potential thing I'm seeing while researching PSUs is that some have over current protection at 30A (even +1kW PSUs, having multiple 12V rails), so a 500W CPU on 12V potentially needs 41A, that would explain exactly the PSU behavior. Looks like the answer is a single rail PSU?
Appreciate any help since this looks like a money hole without an end.
1. the ES CPU doesn't support 6400 speeds or
2. because M321R8GA0EB2-CCPKF (which I'm using) is not listed, but e.g. M321R8GA0EB2-CCPKC is (last 3 letters are different) in the QVL. A strange thing is in BMC under DIMM inventory it sees 96GB installed (but I use 1x 64GB stick) and 6x 16GB (DIMM_P0_A0 till DIMM_P0_F0). ChatGPT says that sometimes this can happen even with most recent BMC and BIOS versions, so I don't think much of it but plan to test with a DDR5-5600 stick.
A whole different problem is running 2 CPUs: if I turn on the PSU, the fans spin briefly and then the PSU makes a sound, like the over current protection would kick in. Now I don't know how much power everything sucks at the boot, but would expect a 1650W PSU (Power LC1650 ) could handle it (also tested on a 1700W PSU (Enermax 1700) with same result). So I wonder what kind of PSUs are people using if normal ATX PSUs get triggered by their over current protection circuits (if that's the case).
One potential thing I'm seeing while researching PSUs is that some have over current protection at 30A (even +1kW PSUs, having multiple 12V rails), so a 500W CPU on 12V potentially needs 41A, that would explain exactly the PSU behavior. Looks like the answer is a single rail PSU?
Appreciate any help since this looks like a money hole without an end.
Last edited:
So, I might have to read your post twice but
did you use the BMC to update to the latest bios?
Then only have in the cpu 0 and 2 socks of ram?
you can actually get to the bios through the BMC to set the memory if that is the concern.
then then using BMC, initiate a reboot
allow the chipset initializing, about 9 minutes worth
wait till final testing on bios screen because you would not have a drive in yet
then power off, add second cpu, and all the ram
power up
then on that bios cycle, turn off, add drive.
As for power:
there are three 6 pin, two 8 pin, and the 24 pin
Some notes, the BMC inventory it items may be wrong till the first few successful boots take place
if you’re power supply is clicking or making bad noises, then the issue is electrical but specifically could be the psu. OR cables
make sure to use the cables from the power supply, pin outs and populated pins can very between manufacturer
example the 6 pin may only have 4 pins in the plug on one brand, 5 in another brand, and 6 in another
lastly
Make sure your cpu socket has all pins even with it any reflections that look astray in the field of gold per aocket
screw down the cpu all the way using even swapping back forth of a screw driver from screw to screw, hat all the screw GENTLY bottom out and then back off each screw 1 full turn
If nothing else works
it’s a psu issue. / cables
Socket seating
And ram as last
when testing only use two stick of ram and limit yourself to socket 0
Power : I doubt, without any video cards in, that the set up is using more than 1100w at power on so you have plenty of power
did you use the BMC to update to the latest bios?
Then only have in the cpu 0 and 2 socks of ram?
you can actually get to the bios through the BMC to set the memory if that is the concern.
then then using BMC, initiate a reboot
allow the chipset initializing, about 9 minutes worth
wait till final testing on bios screen because you would not have a drive in yet
then power off, add second cpu, and all the ram
power up
then on that bios cycle, turn off, add drive.
As for power:
there are three 6 pin, two 8 pin, and the 24 pin
Some notes, the BMC inventory it items may be wrong till the first few successful boots take place
if you’re power supply is clicking or making bad noises, then the issue is electrical but specifically could be the psu. OR cables
make sure to use the cables from the power supply, pin outs and populated pins can very between manufacturer
example the 6 pin may only have 4 pins in the plug on one brand, 5 in another brand, and 6 in another
lastly
Make sure your cpu socket has all pins even with it any reflections that look astray in the field of gold per aocket
screw down the cpu all the way using even swapping back forth of a screw driver from screw to screw, hat all the screw GENTLY bottom out and then back off each screw 1 full turn
If nothing else works
it’s a psu issue. / cables
Socket seating
And ram as last
when testing only use two stick of ram and limit yourself to socket 0
Power : I doubt, without any video cards in, that the set up is using more than 1100w at power on so you have plenty of power
Last edited:
I would try different RAM. If you got to the chipset initializing screen and it stayed there for hours it sounds like the RAM. I had a similar issue. Had to RMA the RAM I had. New RAM worked fine.
Also, seating the CPU in the socket. Make sure you did it correctly. Swapping CPUs around on mine and I had boot issues that ended up being a bad seat job. I don't remember if it got to the chipset screen though.
Also, seating the CPU in the socket. Make sure you did it correctly. Swapping CPUs around on mine and I had boot issues that ended up being a bad seat job. I don't remember if it got to the chipset screen though.