SuperMicro DIMM Voltage difference

[Ewar]

Hi All,

I recently purchased a Supermicro X8DTL-3F fitted with the below hardware but I've noticed some DIMM voltage discrepancies as per my attached screenshot.

The biggest thing is that CPU1DIMM voltage is 1.8+ and CPU2DIMM is 1.4+. The memory installed is 1.5V memory so I'm confused as to why the bank for CPU1 is volted so high when CPU2 voltage is much closer to the rated 1.5V. Also, DIMM voltages are not visible in the BIOS or IPMI at all, they can only be viewed in software like HWMonitor.

Another weird thing is I spun up an old SSD with an expired trial of Server 2012 R2. When Server 2012 R2 trial has expired it causes the computer to shutdown every 60 minutes. When I try to turn it back on after the forced shutdown I can't boot it up as it does the Memory issue beep code - 5 short beeps and 1 long beep. I then have to pull the power and wait a few seconds before it all works again. Supermicro support have told me my memory is wrong but it only happens after this.

When I got the board the IOH/northbridge kept overheating so i've got a 140MM hooked up pointing at it which is keeping it much cooler but I'm wondering if the overvolting on the DIMM is causing it to overheat?

What I've tried:
- Reset BIOS to optimized defaults
- Updated BIOS to latest version
- Updated IPMI to latest version
- Disabled some CPU speed features (now reset back to default)
- Swapped the memory around between banks. CPU1DIMM voltage stays as 1.8 so its not automatic RAM specific

Supermicro support have informed me that this is normal but I'm surprised to see it. Chances are this is in fact how it works or the voltage sensor is screwed so I may have to test it physically.

Installed hardware:
- 2x Xeon x5550
- 6x 4GB Samsung PC3-10600R 1333Mhz DDr3 EEC Registered 1.5v
- 1500w Silverstone Strider PSU - Tested with PSU tester before fitting
- Only 1 SSD connected at the moment

Thanks,
Ed

 

[OBasel]

Are you having issues with your system being stable? Prob not something I'd be concerned about.

 

[Ewar]

Thanks for the reply OBasel.

Initially, I was. I would get IOH overheat after about 20 minutes and the system would shut down. I'm running it in a PC tower case with normal PC fans so I just chalked it down to bad airflow and applied a 140MM fan on the side of my case blowing onto the northbridge and applied new thermal compound to NB and both CPUs. This is what initially made me start looking deeper as I was surprised it the NB was overheating so quickly.

 

[markarr]

Those generations of IOH run really warm and without enough airflow they will overheat and cause your shutdowns. They were not designed to be run in a normal case they were for server cases with higher airflow.