Article View: comp.arch
Article #10967Re: Lisp performance (on Sparc SS2, SS10-30, HP720)
From: Guy Harris
Date: Mon, 16 Nov 1992 18:44
Date: Mon, 16 Nov 1992 18:44
14 lines
1321 bytes
1321 bytes
>From the SpecInt92 results, one should expect a performance increase >of about a factor of 2 when switching from a SS2 to either s SS10-30 >or HP720. Unfortunately, on both machines turned out to but just >30% faster than a SS2 when running our system (for which ps NEVER >shows a resident set under 3-4 MB, not even on an 8MB machine). Just out of curiosity, does the HP machine have a write-through or a write-back cache? I've seen some cases where cache-buster programs run better on machines with a write-through cache than on machines with a write-back cache; your LISP interpreter+application may be such a program. The SS2 has a write-through cache, and the SS10 has, I think, a write-back data cache. (The SS10-30's caches are, when combined, smaller than the SS2's cache - 20K Icache + 16K Dcache, vs. 64K combined cache; how large is the HP's cache?)
Technical Details
Message-ID:
<15506@auspex-gw.auspex.com>Path:
arc1-ca.pugleaf.net!sparky!uunet!auspex-gw!guyReferences:
<1e824rINNlpu@iraul1.ira.uka.de>Show Full Headers
Complete Article Headers
Path: arc1-ca.pugleaf.net!sparky!uunet!auspex-gw!guy From: guy@Auspex.COM (Guy Harris) Newsgroups: comp.arch,comp.benchmarks Subject: Re: Lisp performance (on Sparc SS2, SS10-30, HP720) Message-ID: <15506@auspex-gw.auspex.com> Date: 16 Nov 92 18:44:51 GMT References: <1e824rINNlpu@iraul1.ira.uka.de> Sender: news@auspex-gw.auspex.com Followup-To: comp.arch Organization: Auspex Systems, Santa Clara Lines: 14 Nntp-Posting-Host: auspex.auspex.com