X10SRA + E5-1650v3 + ECC UDIMMs - POST code 67

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

DavidRa

Infrastructure Architect
Aug 3, 2015
322
144
43
Central Coast of NSW
www.pdconsec.net
Trying to track down why my replacement board, an X10SRA, won't complete POST with a CPU and RAM that used to be fine. The previous board was an Asrock X99WS which underwent some form of rapid decompressive failure - something went bang and I can't see which component failed, though I expect whatever it was ended up as smoke.

So I've plugged in the same CPU (Xeon E5-1650v3) and memory (4x 8GB ECC UDIMMs) - the new board powers up, but the POST code display stops at 67 and the board won't boot. If I remove the memory completely, I get the "no RAM" beeps and code 53.

The manual seems to say that UDIMMs are only supported by the i7 processors - but I'm not sure if that's the cause of the failure here. The description for code 67 is wonderfully useless, "CPU DXE initialization (CPU module specific)" so I'm otherwise stumped.

Edit: The only device connected is the video card and USB keyboard. No drives, USB mouse, other devices etc.

Any suggestions?
 
Last edited:

RageBone

Active Member
Jul 11, 2017
611
156
43
well, sadly not really.
53 is the usual No Ram error as expected without Ram, and 67 is for sure beyond the B0 to BF Ram init and training.
I have heard that USB incompatibility can cause weird stuff on post, how about you remove everything even the GPU considering that the board should do fine with just the cpu and a single stick. IF it doesn't, the bang probably killed something.
I kinda doubt its the RAM but the Board could also be kinda faulty.

Is the 1650 an ES ?
I once had the X99 twin of that board and the Bios was horrendous and i think SM patched out ES support on a later Bios.

Depending on your location, i might be able to test the hardware for you if you like.
 

DavidRa

Infrastructure Architect
Aug 3, 2015
322
144
43
Central Coast of NSW
www.pdconsec.net
No, as far as I know it's a legitimate production E5, not an ES. There's no graphics on board - I'm not sure I'd expect it to post without, but it's easy enough to try.

I'm in AU on the Central Coast of NSW, so far away from everyone :D

Edit: Nope. Same code without GPU, without a keyboard, and with only 2 sticks (any two). I wonder if I did manage to fry the CPU somehow!?
 

RageBone

Active Member
Jul 11, 2017
611
156
43
yea, that's far away from everything.

Looks like the CPU got fried somehow. Not 100% sure, could still be the board but yea....

Do you still have warranty?
 

DavidRa

Infrastructure Architect
Aug 3, 2015
322
144
43
Central Coast of NSW
www.pdconsec.net
yea, that's far away from everything.

Looks like the CPU got fried somehow. Not 100% sure, could still be the board but yea....

Do you still have warranty?
No it's been built for at least 2 years if not longer. The board is E-Bay new (this stuff is not really available new, or at reasonable prices in AU). I may grab a spare Precision with the same CPU from a mate and see if I can component-swap to identify.

Thanks anyway though it's very much appreciated.
 

itronin

Well-Known Member
Nov 24, 2018
1,178
759
113
Denver, Colorado
so I just had a similar sounding problem with an eBay X10SRL-F and known good memory and CPU.
I reseated the CPU but that did not fix it. (if you have not you might give this step a shot)
I then backed off the heatsink screws 1/4 turn on each corner. That actually solved it for me.
Not suggesting that is your problem but am offering some ideas.

though given the combustive board failure you suffered the CPU may well have gone with the motherboard (or vice-versa I suppose?)
 

nthu9280

Well-Known Member
Feb 3, 2016
1,627
497
83
San Antonio, TX
just some ideas ...
Not sure if you have DDR3 or DDR4. Reason I ask is that some X99 boards do work with DDR3. This board doest. If you have even a single stick of Reg-DDR4, try that.
Not sure your board has IPMI ( IIRC, they have two SKUs for this board). If you have IPMI, you can try flashing to the latest BIOS and FW.
 

DavidRa

Infrastructure Architect
Aug 3, 2015
322
144
43
Central Coast of NSW
www.pdconsec.net
DDR4 ECC, but unbuffered, not registered. It's the first thing I'll swap out with stuff from the Precision because it's easy and I'm lazy (swapping the CPU is annoying and slow), and I'm _technically_ unsupported (which could mean "if it works fine, but if it screws up you're on your own" or "bend over and spread them, go buy RAM").

It's the non -F version - no IPMI. Workstation, so I didn't need it.

Tried backing out the HSF screws - no luck there either.
 

kapone

Well-Known Member
May 23, 2015
1,092
640
113
Well....in all fairness, you ARE going explicitly against the specifications.

According to Supermicro: UDIMM will be only available for Core i7 processors

The description of the POST code is something that probably less than a dozen people in the world actually understand. It is wonderfully ambiguous.

At this point, my suggestion would be...process of elimination. That's all you really have left. I do realize that "down under" things are tough to procure, but you're kinda stuck. Beg/Borrow/Steal an i7 and test.
 
  • Like
Reactions: Tha_14

DavidRa

Infrastructure Architect
Aug 3, 2015
322
144
43
Central Coast of NSW
www.pdconsec.net
Honestly for $100 or so, this is worth a bid anyway

https://www.ebay.com.au/itm/174153454315

It's not the RAM. Plugged in some loaner RDIMMs from a working Precision T5810 (it's the same platform and CPU) and same error code. Will switch CPUs next...

Edit: OK and it's not the CPU either. Swapped BOTH into the new board, and same code. SuperMicro case lodged too, this is annoying.
 
Last edited:

iGene

Member
Jun 15, 2014
68
21
8
Taiwan
Honestly for $100 or so, this is worth a bid anyway

https://www.ebay.com.au/itm/174153454315

It's not the RAM. Plugged in some loaner RDIMMs from a working Precision T5810 (it's the same platform and CPU) and same error code. Will switch CPUs next...

Edit: OK and it's not the CPU either. Swapped BOTH into the new board, and same code. SuperMicro case lodged too, this is annoying.
Have you found the root cause of this issue? I've also encounter this issue occasionally and haven't found a good solution yet.