Article View: comp.arch
Article #12629Re: % in hash table
From: Kurt Shoens
Date: Wed, 11 Oct 1995 04:04
Date: Wed, 11 Oct 1995 04:04
13 lines
1051 bytes
1051 bytes
djb@silverton.berkeley.edu (D. J. Bernstein) writes: >Hmmm. Do you consider data compression a ``real application''? Absolutely, and if you have numbers, please post them. I have no idea how important the difference between & and % would be in the hash table of a compression algorithm. I'm sure it would be more effective for compression than for, say, gcc. For a largish source file I tried this on (large to get over the startup overhead for gcc), I predict that using & over % would save 0.4% of the compilation time on HP PA-RISC. -- Kurt Shoens
Technical Details
Message-ID:
<45ffol$q2u@news.dtc.hp.com>Path:
arc1-ca.pugleaf.net!news.dtc.hp.com!shoensReferences:
<43tqr6$3au@mikasa.iol.it> <CLIFFC.95Oct5082046@ami.sps.mot.com> <4516lr$t88@news.dtc.hp.com> <1995Oct1102.41.13.1665@silverton.berkeley.edu>Show Full Headers
Complete Article Headers
Path: arc1-ca.pugleaf.net!news.dtc.hp.com!shoens From: shoens@hpl.hp.com (Kurt Shoens) Newsgroups: comp.arch,comp.arch.arithmetic,comp.lang.c Subject: Re: % in hash table Date: 11 Oct 1995 04:04:37 GMT Organization: Hewlett-Packard Message-ID: <45ffol$q2u@news.dtc.hp.com> References: <43tqr6$3au@mikasa.iol.it> <CLIFFC.95Oct5082046@ami.sps.mot.com> <4516lr$t88@news.dtc.hp.com> <1995Oct1102.41.13.1665@silverton.berkeley.edu> Reply-To: shoens@hpl.hp.com NNTP-Posting-Host: kasmo.hpl.hp.com