GIGABYTE MB10 Datto Motherbaord Xeon D-1521 with I/O Shield - $130 + shipping

Marsh

Moderator
May 12, 2013
2,275
1,089
113
I have 2 Gigabyte MB10 boards, I took the ram sticks out of the Gigabyte board to run tests.
I am positive the ram sticks worked for years in the Gigabyte MB10 boards.
 

Evan

Well-Known Member
Jan 6, 2016
3,071
512
113
Wow, for the guys that purchased multiple boards can you imagine why so many bad ? Could it be the seller purchase bulk random stuff in an auction and these boards are all warrently returns where some are actually broken and some just returned but not ?
 

Marsh

Moderator
May 12, 2013
2,275
1,089
113
In the end, seller promptly issued refund , no hassle or question.
All is well.

Besides, I have no idea why I brought 6 boards, I don't need any more Xeon-d board.
It worked out better than buyer remorse. I spent the money on 8 x 8TB disks from Best Buy.
 
  • Like
Reactions: RageBone and alka

acquacow

Well-Known Member
Feb 15, 2017
567
294
63
38
Wow, for the guys that purchased multiple boards can you imagine why so many bad ? Could it be the seller purchase bulk random stuff in an auction and these boards are all warrently returns where some are actually broken and some just returned but not ?
Well, in the case of the system log of the board I received, there were tons of over-temp messages, so I'm assuming these sat somewhere with minimal cooling and/or the fans on the cases failed and they just cooked.
 

ru me

New Member
Jun 2, 2018
18
8
3
Too many broken boards. Not really tested clean pulls as promised. I wasted half a day on this. Well I guess you get what you pay for, in this case nothing.
 

BLinux

cat lover server enthusiast
Jul 7, 2016
2,519
964
113
artofserver.com
this is concerning... although i didn't buy this board, i found a good deal on some other boards that i did buy... now, i'm dreading when they arrive...
 

ru me

New Member
Jun 2, 2018
18
8
3
I am glad that I was not alone in this. In my experience it takes a lot to kill a motherboard during handling. So a clean pull should almost always work. This one was even claimed to be tested. The seller did waste a lot of peoples time. Not cool.
In this case packaging was good but another pet peeve of mine are the sloppy packers. Often I can get a price reduction for the dents. However, all the banging my stuff must have gone through is only acceptable in a hobby situation due to potential hidden damage.
 

PnoT

Active Member
Mar 1, 2015
597
132
43
Texas
this is concerning... although i didn't buy this board, i found a good deal on some other boards that i did buy... now, i'm dreading when they arrive...
Did you get in on the 1541 models from the same seller?
 

PnoT

Active Member
Mar 1, 2015
597
132
43
Texas
I guess you and I are in the same boat. My D1541 motherboard should be here Monday.
A friend of mine went in and we got 8 of the 1541 boards... let's hope they don't turn out like this set as it's utterly ridiculous for a board to not boot on unboxing.
 

mmo

Active Member
Sep 17, 2016
464
242
43
40
man, we need to have a thread to blacklist certain eBay sellers for STH members...
I think it's a good idea to create a list of eBay sellers that STH member should take cautions before buying from them.
 

RageBone

Active Member
Jul 11, 2017
291
77
28
In case anyone has a broken one and is willing to get rid of it for cheap, I'd be interested.
EU by the way.

The standards to check, just to repeat them once more.
1: power on without ram, to know how that behaves.
2: power on with ram.
if Still the same behavior as in case 1, it is either a problem with the RAM, or not the RAM at all.
3: test with one stick and iterate through all slots. CPUs can have bad / dead channels, boards can have damage to ram-traces, some platforms have primary and secondary ram slots.
4: check the ipmi, those Voltage critical messages from page 2 can be a critical hardware defect.
5: Test with a different power-supply. As someone mentioned earlier, some boards and PSUs are picky and are working, but not with and only exactly that board.

6: visually inspect the board for ripped of components, and scratches. Especially around mounting holes.

If you happen to have a TPM Debugg Postcode card, those are damn helpful for such debugging, use it and remember the codes of each step.
If its is stuck on FF, 00, 0D it is at least for now, dead.
bX early on is Ram and Ram training.
00, no cpu, cpu dead or broken board with defective VRMs.
FF is, well a FU**en FAILURE. Some Boards don't reset Port 80 in the liking of the debug cards, making them show FF sometimes.
0D is something i'm still trying to figure out.
 

acquacow

Well-Known Member
Feb 15, 2017
567
294
63
38
Mine hangs at 06 when it's trying to init the BMC, but I couldn't find any lookup table for what 06 means.
 

acquacow

Well-Known Member
Feb 15, 2017
567
294
63
38
06 is very early in the post process.
Had a Supermicro that hang at 04. The cause was a dead oscilator for the BMC.
In theory it shouldn't wait for it, but it did.

https://www.supermicro.com/manuals/other/AMI_BIOS_POST_Codes_for_Grantley_Motherboards.pdf
states micro code loading.
I suggest a bios flash through the ipmi.

EDIT, besides the general debuging.
For instance, does it hang at 06 without Ram?
It hangs at 06 no matter what I do. I already reflashed the bios and the BMC. No change in behavior.
 

sfbayzfs

Active Member
May 6, 2015
247
104
43
SF Bay area
@acquacow - Not sure if there is an option disable BMC on this board. I used have a X9DRD board with borked IPMI. I think I had to disable IPMI to finish posting.
Once I had an X9SCM-F or X9SCL-F with a bad BMC - it wouldn't boot fully, and the VGA output was all messed up, but using the BMC disable jumper, the VGA out was a lot better, and the board worked great thereafter. (Minus the BMC features of course obviously)
 

acquacow

Well-Known Member
Feb 15, 2017
567
294
63
38
I don't have any jumpers labeled as such, but there are some unlabeled jumpers I could experiment with. I could unlock and edit the bios and see if there are bios options for disabling the BMC as well.