H11DSi sudden restart crashing

[norcimo]

Greetings

I have a NEw H11DSi (rev 2) that randomly restart by itself, causing issues with the operating system as it crashes. it happens not under load.

im using 2 7742 ES 100-000000053-04, with 4 sticks of 16gb DDR4 3200,, installed on slots P1Dimma1, p1Dimmb1, and p2Dimma1 , p2Dimmb1 64 gig total.

I tested the memory with a USB boot disk memory tester: OK.

I also tested using cinebench I (when it did not restart) and it also passed good with 30541 mark.

Im was using Windows 10 LTSC 64bit

Im trying today Windows Server 2019, but that also crashed on the first try. and currently attempting again.

Any thoughts?

 

[norcimo]

Update: I could not install Windows Server 2019 till I went into the bios and ramped down the CORES to 2x2. Weird.

 

[ari2asem]

in bios:
- disable SMT
- install memory according the manual, because you dont do that
- disable (or enable IOMMU). just play with it. sometimes disable helps, sometimes enable helps.
- try to use sata port (sata ssd/hdd) as boot os disk. i had issues with m2.nvme ssd's. after many try-error attemps i chose to go with 2.5inch sata ssd.
- try to install win 10 enterprise

 

[norcimo]

in bios:
- disable SMT
- install memory according the manual, because you dont do that
- disable (or enable IOMMU). just play with it. sometimes disable helps, sometimes enable helps.
- try to use sata port (sata ssd/hdd) as boot os disk. i had issues with m2.nvme ssd's. after many try-error attemps i chose to go with 2.5inch sata ssd.
- try to install win 10 enterprise

"install memory according the manual, because you dont do that " ok the manual I can't understand it well. I have 4 sticks 16GB each ddr4-3200... what is your recommendation for placement according to the manual?

 

[i386]

I don't know how well engineering sample cpus go with a rev. 2 board with a current firmware/bios.

 

[norcimo]

its been a disaster! either the chips are bad or a bad motherboard.
I'm able to install windows 10 ENT , ltsc, and server 2019. on NVME or SSD, dont matter .. i just randomly crashes

 

[norcimo]

Maybe I need more memory. I'm going to reduce the CCD settings and see what happens as I wait for the sticks to show up.

 

[TXAG26]

As mentioned above, read the manual as to which slots your 4 sticks of ram need to populate. Supermicro boards are very particular about that.

 

[norcimo]

As mentioned above, read the manual as to which slots your 4 sticks of ram need to populate. Supermicro boards are very particular about that.

Can you help me identify that? I cant understand that memory table in t
he manual

 

[TXAG26]

If you have a total of 4 sticks of DDR4 for the board, it looks like they need to be installed in the following slots:
CPU P1: C1 & D1
CPU P2: C1 & D1

See pages 32-33 of the manual for the board diagrams and labels.

 

[norcimo]

ok, i was not sure.. will try , thanks

 

[norcimo]

well. it seems to last longer. Good. I also got in four more sticks. So I occupied the rest. Seems a bit more stable, only one crash. I may need to re-install the OPS again.

 

[norcimo]

still doing it. I think the MB is messed up. I'm going to request a new one.

 

[norcimo]

Question? could it be the RAM clock speed? these are 3200, can the 7742 ES 100-000000053-04 have an issue?

 

[norcimo]

Another Symptom, with more memory, the longer it takes to reset, again under no load.

 

[ari2asem]

i got the same hardware as you have (h11dsi rev 2 and dual 7742 with 04 at the end).
my bios is 2.0b.
i have 8 dimms of sk hynix and 8 dimms of crucial, all 3200mhz speed and installed in mixed way (cpu1 has both, sk hynix and crucial, cpu2 has also both, sk hynix and crucial)

when i very very first time booted my system (thus before installing windows 10), i only saw 15 dimms. reseating the cpu solved my memory problem.

try these 3 things:
- update your bios to the next version (thus not immediatly to the newest bios version)
- reseat your cpu's
- set all to default in bios and start again with try-error attempts

 

[norcimo]

I reseated both CPUs today. The bios that it came is version 2.1 maybe I should go back one version just in case. this is rev 2 this motherboard so I should be able to go back one version without a problem and try.

 

[norcimo]

i got the same hardware as you have (h11dsi rev 2 and dual 7742 with 04 at the end).
my bios is 2.0b.
i have 8 dimms of sk hynix and 8 dimms of crucial, all 3200mhz speed and installed in mixed way (cpu1 has both, sk hynix and crucial, cpu2 has also both, sk hynix and crucial)

when i very very first time booted my system (thus before installing windows 10), i only saw 15 dimms. reseating the cpu solved my memory problem.

try these 3 things:
- update your bios to the next version (thus not immediatly to the newest bios version)
- reseat your cpu's
- set all to default in bios and start again with try-error attempts

Where can I get your bios? 2.0b.

 

[ari2asem]

bios version 2.0B for h11dsi rev.2 board

Gofile

 

[norcimo]

bios version 2.0B for h11dsi rev.2 board

Gofile

Thank you. GOing to try it out..

right now im going stress test the CPUS with PRIME 95 and see what happens