Path: utzoo!attcan!uunet!cs.utexas.edu!rice!erato.rice.edu!cliffc
From: cliffc@erato.rice.edu (Cliff Click)
Newsgroups: comp.arch
Subject: Re: Compiler Costs
Message-ID: <9730@brazos.Rice.edu>
Date: 9 Jul 90 13:33:53 GMT
References: <1797@apctrc.UUCP> <373@e2big.mko.dec.com> <2319@l.cc.purdue.edu>
Sender: root@rice.edu
Organization: Rice University, Houston
Lines: 21

In article <2319@l.cc.purdue.edu> cik@l.cc.purdue.edu (Herman Rubin) writes:
>But one is unlikely to produce an artist by having the apprentice color within
>the lines.  
>-- 
>Herman Rubin, Dept. of Statistics, Purdue Univ., West Lafayette IN47907
>Phone: (317)494-6054
>hrubin@l.cc.purdue.edu (Internet, bitnet)	{purdue,pur-ee}!l.cc!cik(UUCP)

For once I agree with you!

Looking at what some compilers produce (say, at least 6 commerical ones) I can
get a factor of 2 speedup by doing "stupid" optimizations, that I would expect
the compiler to get on it's own.  To get better than that I cannot color 
within the compiler's lines, but generally I need to restructure how things
are solved - not nessecarily a change of algorithms but a change of *structure*.

Cliff Assembly Hacker Click

--
Cliff Click                
cliffc@owlnet.rice.edu