I have the exact same issue.

I am still not certain I found the issue, but I did catch some "major event" a couple of weeks ago and found that the OS was writing thousands of entries to the Event Log (like many per second). When I googled the error, it appeared to be an ECC/memory error that I guess it was trying to self-correct. (That took me a relatively long time to figure out, because the issue was consuming a ton of resources.) I did not let it get any further, and did a shutdown - so I don't know if it would have actually fully frozen.

FWIW: I can not get memtest86+ to run on the machine. It just crashes soon after starting - seems to be something with Dell BIOS that won't run it. I did run the BIOS memory test for a few hours, and it didn't report anything so very strange.

I ordered new memory and swapped it out completely. I have not had the same issue since, but that's only a few weeks so far.

I also suspect this could be a RAM issue. The diagnostics show no warnings/alerts on Memory.
This T7600 has 128 GB RAM (16*8), It's my personal labbing machine, so, I can't replace the whole RAM at once, Guess I'll have to remove one by one until I hit the faulty one.

The problem with trying to replace one stick at a time is you'll almost certainly end-up with mismatched RAM and who knows if that'll cause different/additional issues.  This is a sticky situation (troubleshooting RAM when there's no obvious culprit).

My recommendation would be to, instead, remove 1/2 your RAM for some time...I assume you can tolerate running with 64GB for your workload.  See if you still encounter the issue.  If you DO, switch the set of 64GB and try again.  (Really whether you do or don't get the error...)

You could start to isolate the stick that's going bad that way (you'll have to repeat the process in an organized and methodical way).

If you don't get any conclusive results...well, more mystery, unfortunately.

Best of Luck.

1,300w is output D.C.  Check the label, either on PSU or tower for A.C. input and get a PSU that can exceed that and run PC as long as you want.  A PSU might be rated as 10 amp for 15 min., for example, and so on.  Also remember you won't always be drawing full power.

I couldn't find any documemtation to be any more specific, let alone immediately find an image of a label indicating input amps.

What I have ran into in the not so distance past is make sure the UPS puts out a true sine wave, not just simulated.

So far the issue still persists.

I've changed RAM, eventually same behavior, now it's not lasting even a day turned on.

Ran a diagnostics and still no alarms.

What is the known behavior of a faulty PSU?

I'm about to throw this thing away.

Ahhh! This was the same problem I had been through on my Dell T3610 for almost a month and finding a solution. I took it back to the shop vendor I bought this PC from and he tried almost all ram modules he had but the error did not go away. I am certain that it is a RAM compatibility issue or maybe a RAM slot issue. When I changed all of my 4 RAM modules from 2 channels to all 4 channels (1 module on each) the Error was gone. I still did not know the reason but just ignored it as it might be due to a bad RAM slot. I would suggest that before changing the RAM slots check a non-ECC RAM. I wished I could have checked that as well. I used the Memtest86 to check the RAM.

Hey @aweber1nj , any Update? Did you find the solution?