How do I correctly specify the required channel, username, and password for IPMICFG?
IPMICFG.efi -a ##:##:##:##:##:##
Failed to set the MAC Address. Completion Code=82h
Last edited:
GIGABYTE MZ31-AR0 REV 1.1 with Epyc Rome / Milan
- Thread starter Arcanum
- Start date
-
- Tags
- bios epyc motherboard
Notice: Page may contain affiliate links for which we may earn a small commission through services like Amazon Affiliates or Skimlinks.
IPMICFG.efi -a ##:##:##:##:##:##
Failed to set the MAC Address. Completion Code=82h
if not work you may need to use eeupdate instead.
or you can check the binary for default MAC and edit it with a Hex editor.
Trying to update a Rev 1.1 to support Epyc Gen2 (Rome) as well. Tried flashing via BCM (RBU versions that were shared here that supposedly supported it), but I get a failure to upload - after that, the system refuses to post (wonder if it actually succeeded and I should switch the CPU for the gen2 one). For the guys where that has worked out, do you get a success message in that case?
I am owner of the 2 MZ31-AR0 rev.1.1 motherboards, so, here is my 5 cents:
1. EPYC 7371 are happy with R40_F23 BIOS from the rev. 2.0 motherboard - obviously, changing SPI flash IC from 128Mbit to the 256Mbit is required.
2. EPYC 7F32 with rev.2.0 BIOS R40_F23 or rev.3.0 BIOS M15_R40 wouldn't POST, it randomly freezes on the "c0", "d0" or even "00" POST codes.
3. I tested 3 motherboards overall - one by "full" rev 2-3 BIOS'es, two others by "halves" of rev.2-3 BIOS'es. Same story.
4. Serial number printed on the label on the ATX connector had datacode - first 4 digits: year and the week of the year. So, "1743", "1803" and "1819" is no go.
5. There is evidence in revision history of the rev. 2.0 "Rome" part of the BIOS about updating VRM firmware:
"T06
1. Update VR(IR) rom version(20190411)"
and
"T11
2. Update VR(IR) version to A02(8144)"
Obviously, there is no way to find such "T" "test" BIOS'es somewhere in the internet, unfortunately.
However, by gathering evidence and good will it is possible to do something:
a. Certainly, from some point MZ31-AR0 rev. 1.1 motherboards starting to use updated VRM firmware - obviously, there is no practical reasons to program separately VRMs of the otherwise practically the same rev.1.1, rev. 2.0 and rev. 3.0 motherboards.
b. Such updated firmware is no available freely in the internet.
c. However, it is possible to read and program VRM controller by software and USB dongle.
So, I would ask people, if they had rev. 1.1 motherboards running "OK" EPYC's Rome and Milan:
I. At least, please provide your SN on the label on the ATX connector.
At least first 4 numbers if you don't want provide the full serial number for whatever reason.
II. Somebody who wants to be a volunteer - for dumping VRM firmware by USB dongle and software.
I would start that (dumping and programming VRM) in the near future by myself - first reading and than flashing VRM firmware back - for being sure it works as intended (reading and flashing VRM controller).
Feeling free to ask questions or explanations.
Thanks!
1. EPYC 7371 are happy with R40_F23 BIOS from the rev. 2.0 motherboard - obviously, changing SPI flash IC from 128Mbit to the 256Mbit is required.
2. EPYC 7F32 with rev.2.0 BIOS R40_F23 or rev.3.0 BIOS M15_R40 wouldn't POST, it randomly freezes on the "c0", "d0" or even "00" POST codes.
3. I tested 3 motherboards overall - one by "full" rev 2-3 BIOS'es, two others by "halves" of rev.2-3 BIOS'es. Same story.
4. Serial number printed on the label on the ATX connector had datacode - first 4 digits: year and the week of the year. So, "1743", "1803" and "1819" is no go.
5. There is evidence in revision history of the rev. 2.0 "Rome" part of the BIOS about updating VRM firmware:
"T06
1. Update VR(IR) rom version(20190411)"
and
"T11
2. Update VR(IR) version to A02(8144)"
Obviously, there is no way to find such "T" "test" BIOS'es somewhere in the internet, unfortunately.
However, by gathering evidence and good will it is possible to do something:
a. Certainly, from some point MZ31-AR0 rev. 1.1 motherboards starting to use updated VRM firmware - obviously, there is no practical reasons to program separately VRMs of the otherwise practically the same rev.1.1, rev. 2.0 and rev. 3.0 motherboards.
b. Such updated firmware is no available freely in the internet.
c. However, it is possible to read and program VRM controller by software and USB dongle.
So, I would ask people, if they had rev. 1.1 motherboards running "OK" EPYC's Rome and Milan:
I. At least, please provide your SN on the label on the ATX connector.
At least first 4 numbers if you don't want provide the full serial number for whatever reason.
II. Somebody who wants to be a volunteer - for dumping VRM firmware by USB dongle and software.
I would start that (dumping and programming VRM) in the near future by myself - first reading and than flashing VRM firmware back - for being sure it works as intended (reading and flashing VRM controller).
Feeling free to ask questions or explanations.
Thanks!
IR is Infineon, which exact controller is on it ?
there are almost 3 or more different gens of controllers need diff. software and Dongle, all need licence(hard but possible to get)
PowerCode
PowIRCenter
etc.
It's IR35201
Here's thread regarding programming it:
ASUS PWM ASP1405I (=IR35201) firmware contents for ROG MAXIMUS XII HERO (WI-FI) - Badcaps
Hi I have an ROG MAXIMUS XII HERO (WI-FI) board with demaged PWM controller ASP1405I, and this is unfortunalely a programmable chip. Without programming it, it won't start to work on the board - have checked this physically with a new unprogrammed chip. I'm looking for information about the...
www.badcaps.net
internal Test/Debug BIOSes. some with good contacts to gigabyte can get them
seems Hetzner 1.1 boards have (link)
ok i have purchased a MZ31-AR0 1.1 (219€ with 10€ code)
there is a HZ sticker on it, so i guess it is a Hetzner board and should have right VR firmware on it.
i found a module including "8144" in the R03 BIOS, it is located in the ROM Hole, normaly where start pictures are stored.
maybe the Update VR file is still included in the BIOS and waits to be executed.
there is a HZ sticker on it, so i guess it is a Hetzner board and should have right VR firmware on it.
i found a module including "8144" in the R03 BIOS, it is located in the ROM Hole, normaly where start pictures are stored.
maybe the Update VR file is still included in the BIOS and waits to be executed.
Is it from eBay store "RAM-KÖNIG"?
BTW, can you tell the SN of the motherboard?
And, where is "HZ" sticker located on it?
Regards VRM firmware module - should be special command for flashing it by BIOS flash update program or by BMC update program, I guess.
Or even specific flash program for that.
BMC flash update program specifically lists 2 SPI flash available and 3 NOR ones during update.
Maybe even USB dongle doesn't needed for VRM read/update.
Will see, currently busy by moving to another flat.
Thank you!
yes,
purchased it, waitung for delivery.
That's funny!
Because, all 3 motherboards I had dealt with are also from that seller!
Two's mine and one of my colleague.
They (seller) must have huge stock of those, definitely there is probability of some motherboards with a updated VRM firmware.
BTW, where you got "R03" BIOS? Can you provide a link or a file?
Thanks!
the VRM file should be included in later BIOS too.
for MZ32-AR0 actual VR version is 8228, and in BIOS i found a file in Ami ROM hole which contains "8228" in same position....
ram-koenig
SN1806
flashed Rome BIOS but 7402P not work.
no HZ sticker(only on ebay listing picture)
i guess they know 7002 is not work.
"Welcome to the club!"
If the VRM firmware module already located by you - it's only matter of the time to discover how to flash it by pure software way, or, trough USB005 dongle and software.
What I'm already discover - that motherboard completely ok with 256Mbit MXIC MX25L25673G and MX25L25645G - however for the 45G you need to enable quad SPI mode configuration bit (73G had that bit hardwired).
Rev. 2.0 last BIOS R40_F23 works ok with EPYC 7371, so, there is no restrictions in the way Naples part initialized and started.
And, because Rome tries to start (not always "00" on POST, usually"d0" and "c0"), CPU aren't on hold or "RESET" - I can switch power on/off by button or trough BMC - it is most likely power issues by VRM.
Which is dependent on VRM firmware.
I hope, that will be figured out in near future.
the VRM firmware exists in Rome/Milan BIOS not Naples BIOS. i need to run Rome CPU first in order to let it auto update the VRM.
But, for start, it requires at least some basic, "Rome compatible" version.
Otherwise, it wouldn't start.
I don't believe that there is some real hardware difference between older and newer Rev. 1.1 boards.
Only factory flashed different VRM firmware versions.
And, if it's too old - doesn't incorporate "Rome tuning" - it wouldn't start.
If you know the way BIOS update program initializes VRM firmware update with "Rome" CPU - maybe there is the way to force it with "Naples" one.
correct: i mean RUN. not just have.
correct: the new VR updates are from rev. 2.x so 1.1 can't have them from factory.
have only the raw register data. need to find the updater.