X11DPi-NT: Cannot boot, upgrade BMC firmware or BIOS

[NablaSquaredG]

Hey,

I've got an X11DPI-NT here with BMC Firmware 1.7.11.
I want to upgrade the BMC Firmware, but once the firmware is uploaded, I either get a white screen (Supermicro frame is there, but no content) or an "500 Internal Error" when I try again.
If I monitor my upload speed I can see that approximately the size of the BMC firmware is actually being sent.

When I try to update the BIOS, it uploads (similar to the BMC firmware), but then I get an "BIOS Upload Failed"

When I try to POST the machine (equipped with 2x QQ9E ES) the fans turn on for ~1 second and then immediately turn off. Health log says "Thermal Trip"


Is the board fried?

 

[NablaSquaredG]

Alright, progress!

I was able to get the board to boot with just 1 CPU to Linux, from there I was able to flash the 01.73.12 via the sum tool.

Now upgrading the BIOS via IPMI... Let's see how this goes.

 

[NablaSquaredG]

BIOS successfully upgraded to 3.5

Now the real issue:
The CPUs (QQ9E ES) run fine alone, but both together won't work.

Any ideas?

 

[Spearfoot]

BIOS successfully upgraded to 3.5

Now the real issue:
The CPUs (QQ9E ES) run fine alone, but both together won't work.

Any ideas?

If 'ES' means 'Engineering Sample', that's probably why they don't work together.

 

[NablaSquaredG]

If 'ES' means 'Engineering Sample', that's probably why they don't work together.

The seller I bought the CPUs from (from this forum) has shown benchmarks with the CPUs running in dual socket setup.

 

[Spearfoot]

The seller I bought the CPUs from (from this forum) has shown benchmarks with the CPUs running in dual socket setup.

Perhaps in a different motherboard?

 

[NablaSquaredG]

Perhaps in a different motherboard?

Nope, X11DPi-NT too...

 

[RageBone]

@NablaSquaredG i have heard of the socket being finicky with pressure and sum such.
Reseating the second cpu might do it.
Another thing to look out for is standoffs in the wrong place.

POST Codes won't help if it powers off immediately.
Please test the second cpu alone to make sure it still works.

The board could be damaged in a way that it kills CPUs.
So if the second one still works alone, that isn't the case.

Bent pins is another thing to look for or any other physical damage of the board like scratches and ripped of parts.

 

[NablaSquaredG]

Reseating the second cpu might do it.

Done a couple of times, no luck.

I have also swapped CPU1 and CPU2 (yes, I had to rotate the carriers on the coolers for that to work), no luck.

I have double checked for the right orientation of the CPU (triangles)

Another thing to look out for is standoffs in the wrong place.

Double checked.
X11DPi-NT is placed in a Supermicro 826 chassis and mounted on 10 standoffs (2 standoffs are beside the board because not needed) -> Makes 12 standoffs. 2 standoffs are beside the board and because the 10 screws of the mobo hold tight, the standoffs must be positioned correctly

Please test the second cpu alone to make sure it still works.

Tested. Both work fine individually.

Bent pins is another thing to look for or any other physical damage of the board like scratches and ripped of parts.

Second Socket looks fine:

http://imgur.com/a/GShDGb5

First one is also 100% good

 

[RageBone]

2x 8Pin EPS? Usually powerplanes are unified but who knows.
Shouldn't cause a poweroff anyway.

I'd say everything points to a faulty board.

 

[NablaSquaredG]

Yep, faulty board.

CPUs work fine on Gigabyte MD71-HB0...