Souper Computer

[jrv at vanzandt.mv.com (James R. Van Zandt)]

Scott Lipcon <slipcon at cs.jhu.edu> writes:

>compare it to a uniprocessor system... "Speedup" is defined as speed of 
>a uniprocessor divided by speed of the parallel machine.  For an N 
>processor machine, the maximum possible speedup is N.  You wont get 
>that in real life.  A simple model is Amdahl's Law, which says the 
>following:
>
>Assume that for a given program, x% can be run only on one processor.  
>Also assume that the rest of the job is entirely parallelizable.  
>Therefore (1 - x)% of the job can be run on all N processors.  The 
>speedup in this case is:
>N / ( 1 + (N - 1)x)

Several years later, another researcher (whose name I have
unfortunately forgotten, but you will probably hear about in your
course) "refuted" Amdahl's Law, in the following sense: Amdahl's Law
is based on solving a fixed size problem faster.  It is often more
useful, and more feasible, to solve a bigger problem in about the same
time.  That is, you can come closer to solving an N times bigger
problem in the same time than you could to solving a fixed size
problem in 1/N the time.

		- Jim Van Zandt
-
Subcription/unsubscription/info requests: send e-mail with subject of
"subscribe", "unsubscribe", or "info" to discuss-request at blu.org