Author: Robert Hyatt
Date: 08:31:21 10/02/03
Go up one level in this thread
On October 02, 2003 at 04:18:20, Daniel Clausen wrote: >On October 01, 2003 at 22:39:55, Jeremiah Penery wrote: > >>On October 01, 2003 at 20:29:35, Ricardo Gibert wrote: >> >>>Here is an example of a formula that always increases as x increases and yet >>>remains bounded: y = 1 - 1/(2**x) >> >>That has nothing to do with Vincent's claims. > >No, but he answered to Bob and his "If efficiency continues to climb, it is >unbounded.". :) > >Sargon My claim is, and always has been, that with N processors, the maximum average speedup is bounded by N. And, in fact, N is a bound that can't be reached except through round-off as in the Cray Blitz article where a speedup was rounded to 1 or 2 places for compactness. However, Vincent has _always_ claimed he could do > 2.0 speedup with 2 processors. If you can go beyond 2 with two processors, what is going to be the limit or bound for the function? Nothing I can see.
This page took 0 seconds to execute
Last modified: Thu, 15 Apr 21 08:11:13 -0700
Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.