OT: Weird Hardware Problem

[Valeri Galtsev]

On 5/20/20 12:41 PM, Tim Daneliuk wrote:
> On 5/20/20 11:07 AM, Valeri Galtsev wrote:
>> Yes, it was I who mentioned electrolytic capacitors.
> 
> Yeah, I am painfully aware of these issues, having started my career as
> an analog circuit designer :)
> 
> What made this one so weird is that the PS that isn't working on the Dell,
> worked just fine on another machine that had _more_ drives in it and an
> old power hungry AMD chip in it.
> 

If you mean AMD CPU, then I can tell you that these CPUs opposed to 
Intel ones are awfully robust. You can literally boil water on AMD CPU, 
and it will keep running without glitches. I had once to have AC filters 
replaced in small server room VAC system which only could be done if all 
AC (air conditioning) is switched off. That took almost two hours during 
which main servers and number crunchers were left running. The 
temperature in the room was 96 F (36 C) - like on a beach in Honolulu. 
Few Intel based boxes got sick (not all if them but significant 
portion). None of AMD CPU based machines failed, they just kept running.

Another factor: older CPUs have higher internal voltages. As you know, 
internal logics, like ALU, run at much lower voltages than those 
supplied to CPU. Though these are not directly related to external 
voltages on CPU pins, still... they quite likely will be more tolerant 
to higher voltage ripple in case of higher internal voltages.

Valeri

> 

-- 
++++++++++++++++++++++++++++++++++++++++
Valeri Galtsev
Sr System Administrator
Department of Astronomy and Astrophysics
Kavli Institute for Cosmological Physics
University of Chicago
Phone: 773-702-4247
++++++++++++++++++++++++++++++++++++++++