Path: utzoo!utgpu!water!watmath!clyde!att-cb!att-ih!pacbell!ames!pioneer!eugene
From: eugene@pioneer.arpa (Eugene N. Miya)
Newsgroups: comp.arch
Subject: Re: Powers of 2, was Re: RISC is a nasty no-no!
Message-ID: <5718@ames.arpa>
Date: 8 Mar 88 17:09:26 GMT
References: <17415@think.UUCP> <1339@alliant.Alliant.COM> <24529@yale-celray.yale.UUCP> <4734@ecsvax.UUCP>
Sender: usenet@ames.arpa
Reply-To: eugene@pioneer.UUCP (Eugene N. Miya)
Organization: NASA Ames Research Center, Moffett Field, Calif.
Lines: 19

In article <4734@ecsvax.UUCP> urjlew@ecsvax.UUCP (Rostyk Lewyckyj) writes:
>I believe that on the CRAY the magic number for bank conflicts
>is something like 48 which is not a power of two.
>On the IBM 3090 16 is a magic number because of the sructure
>of the cache rather than memory bank interleave.

48? Naw, many of us have measured it as powers of 2 (now it depends
on your memory size as to the worst stride).  But, more to the point
I've not measured the factor of 16 on the 3090 (A 200 in my case, and I have
measured conflicits at 128 elements), can you show me and the net some data?
(Not a literature reference, [I have Dorn's TR] but some real data and
a code chunk if you have it).

From the Rock of Ages Home for Retired Hackers:

--eugene miya, NASA Ames Research Center, eugene@ames-aurora.ARPA
  "You trust the `reply' command with all those different mailers out there?"
  "Send mail, avoid follow-ups.  If enough, I'll summarize."
  {uunet,hplabs,hao,ihnp4,decwrl,allegra,tektronix}!ames!aurora!eugene