[Gllug] re: SCSI vs SATA

Nix nix at esperi.org.uk
Sat Jun 12 00:08:08 UTC 2004


On Wed, 9 Jun 2004, Andy Farnsworth muttered drunkenly:
>> Therefore, if the distance the signal has to travel is more than 
>> 88.235 mm, the process will take longer than one cycle of the 
>> processor. That is a little over 3.5 inches and shrinks each and every
> 
>> time the clock speed increases.
> 
> You're forgetting about pipe-lining.  Still, at least on earlier intel
> processors, a multiply did indeed take 30 or 40 cycles and some
> instructions took over 100 cycles!  Also, waiting for memory/io takes
> time too.
> 
> -- /You said --
> 
> Pipelining only works if your system has something to put into the
> pipeline other than "your process" so I gave the simplest example here.

Errr, pipelining has nothing whatever to do with multiprocessing; indeed,
the pipeline generally is emptied when a context switch occurs.

What pipelining *does* do is keep code and data flowing from the RAM
(often more than 88mm signal travel time away) without the CPU needing
to wait for it very often. (Ideally.)

-- 
`If you believe in strong AI, then death is no longer a mystery,
 but merely a lack of adequate backups.' --- Steven McDougall
-- 
Gllug mailing list  -  Gllug at gllug.org.uk
http://lists.gllug.org.uk/mailman/listinfo/gllug




More information about the GLLUG mailing list