[TriLUG] AMD and reliability

Randall Barlow rpbarlow at eos.ncsu.edu
Wed Aug 3 16:51:41 EDT 2005


It makes sense that both AMD and Intel would have similar success rates 
on their chips - these things aren't the sort of thing that typically go 
"bad" once they've made it past the manufacturing stage.  After 
manufacturing, the chip either works or it doesn't, and it is tested in 
the factory with a BIST (Built in Self Test) pin before it goes to 
market to verify against defects.  After that, the only way I can think 
of a chip having problems is heat, which is often times problems with 
the case or cooling systems, and not the chip itself.   Heat usually 
even just causes errors in data and not usually permanent damage, though 
that is possible if the temperature got high enough.  So I wouldn't 
expect reliability differences between AMD and Intel, just performance 
differences since they have different implementations of the ISA...

Randy Barlow

Lee Fickenscher wrote:

> I have to agree with Ken 100%.
> I have found basically zero difference between the reliability of AMD  
> and Intel. Your choice of motherboard and/or memory is much more  
> likely to affect your stability than your choice of processor.
> I generally choose AMD just because I feel that they have a better  
> price/performance ratio. They certainly have a better power/ 
> performance ratio lately with the exception of the Pentium M, which  
> would probably be the only Intel chip I would choose to use over AMD.
> -Lee





More information about the TriLUG mailing list