Re: ugh, moore's "law"

[Lars Clausen <lrclause cs uiuc edu>]

On Fri, 23 Aug 2002, Jason Maiorana wrote:
[...]
> CPU/compenent speeds have been supposedly going up, yet somehow
> computers seem aesthetically slower as the years go by.

At a conference a few years back, one talk included a graph of CPU speed
(in raw MHz) vs. benchmark performance.  The performance/MHz was just about
constant.  What have all those compiler writers been doing all these years?

Two things: Improving usability and keeping up with the CPUs.  Many CPU
changes, notably RISC, assumed that compilers were smart enough to make
good code.  That's extra burden on the compilers.  Also, while Java may be
slow in raw performance, it's fast when you factor in development time --
in most cases.  There's a reason Dia is in C and many scientific programs
are in Fortran.  It's all about the right tools.

Anyway, we're way off topic here.

-Lars

-- 
Lars Clausen (http://shasta.cs.uiuc.edu/~lrclause)| Hårdgrim of Numenor
"I do not agree with a word that you say, but I   |----------------------------
will defend to the death your right to say it."   | Where are we going, and
    --Evelyn Beatrice Hall paraphrasing Voltaire  | what's with the handbasket?