Once in a blue moon, I am seeing a blue screen of death on a shiny new Dell R7610 with a single 1100 Watt Dell-provided power supply on a beefy UPS. BCode is 101 (A clock interrupt was not received...), which some say is caused by under-volting a CPU.

Naturally, I would have to contact Dell support, and their natural reaction would be to replace a motherboard, a power supply, or CPU, or a mixture of the above components.

In synthetic benchmarks, system memory and CPU, as well as graphics memory and CPU perform admirably, staying up for hours and days.

My questions are:

1. Is power supply good enough for the application? Does it provide clean enough power to VRMs on the motherboard?
2. Are VRMs good enough for dual Xeon E5-2665?
3. Does C-states logic work correctly?
4. Is there sufficient current provided to PCIe peripherals, such as disk controllers?

P.S. Recently, I've gone through the ordeal with HP. They were nice and professional about it, but root cause was not established, and the HP machine still is less than 100%, giving me a blue screen of death once in a couple of months.

Here's what quick web-searching turns up: http://www.sevenforums.com/bsod-help-support/35427-win-7-clock-interrupt-bsod-101-error.html#post356791

It appears Dell has addressed the above issue by clocking PCIe bus down to 5GT/sec in A03 BIOS. My disk controllers support PCIe 3.0, meaning that I would have to re-validate stability. Early testing shows improvements.

Further testing shows significant decrease in performance on each of the x16 slots with Dell R7610 with A03 BIOS. But now it's running stable.

HP machine has received a microcode update in September 2013 SUM (July BIOS) that makes it stable.