Path: utzoo!utgpu!jarvis.csri.toronto.edu!cs.utexas.edu!tut.cis.ohio-state.edu!ucbvax!decwrl!shelby!portia!underdog
From: underdog@portia.Stanford.EDU (Dwight Joe)
Newsgroups: comp.arch
Subject: Re: Deja vu?  Architecture book name?
Message-ID: <8861@portia.Stanford.EDU>
Date: 6 Feb 90 02:37:15 GMT
References: <12660001@hpcuha.HP.COM>
Sender: Dwight Joe <underdog@portia.stanford.edu>
Reply-To: underdog@portia.Stanford.EDU (Dwight Joe)
Organization: Gigantor Institute of Applied Science
Lines: 25

In article <12660001@hpcuha.HP.COM> kulkarni@hpcuha.HP.COM (Sandeep Kulkarni) writes:
|I know this was mentioned in this newsgroup some time ago, but I can't seem to
|remember the discussion.  Could someone e-mail me the name of the
|not-yet-released architecture book by Hennessy and Patterson.  I'd like the
|name of the publishers of the book too.

The book is "Computer Architecture:  A Quantitative Approach",
published by Kaufman.

I've read several of the chapters; they are interesting.
However, you'll have to take one assertion in the preface (introduction?)
with a grain of salt.  There, the authors claim that 
computer architecture WILL become a hard science like physics
and chemistry and that this book is a step in that
direction--hence, the reason for the title.  (chuckle) 

I'll admit that there is more quantative material than 
other computer arch. texts.  In fact, the calculations
are reminiscent of molar, molal, and normal calculations in
chemistry.  However, its stretching things a bit far
to say that computer architecture will become a hard science.

In order to qualify for a hard science, there needs
to be a high level of mathematical rigor, which computer
architecture will NEVER have.