From info-hol-request@leopard.cs.byu.edu Tue Nov 29 09:08:23 1994
Received: by leopard.cs.byu.edu
        (1.38.193.4/16.2) id AA13857; Tue, 29 Nov 1994 09:08:23 -0700
Sender: info-hol-request@leopard.cs.byu.edu
Errors-To: info-hol-request@leopard.cs.byu.edu
Precedence: bulk
Received: from swan.cl.cam.ac.uk by leopard.cs.byu.edu with SMTP
        (1.38.193.4/16.2) id AA13851; Tue, 29 Nov 1994 09:08:03 -0700
Return-Path:
Received: from auk.cl.cam.ac.uk (user jrh (rfc931)) by swan.cl.cam.ac.uk
          with SMTP (PP-6.5) to cl; Tue, 29 Nov 1994 16:04:14 +0000
To: info-hol@leopard.cs.byu.edu, qed@mcs.anl.gov
Subject: Re: FWD: NYTimes article on Pentium bug
In-Reply-To: Your message of "Sun, 27 Nov 1994 11:54:31 PST." <9411271954.AA06250@maui.cs.ucla.edu>
Date: Tue, 29 Nov 1994 16:04:03 +0000
From: John Harrison
Message-Id: <"swan.cl.cam.:297130:941129160419"@cl.cam.ac.uk>


I'm curious about whether the problem was in the underlying algorithm used
or its hardware implementation. If I interpret the press report correctly:

| Intel said the error occurred because of an omission in the translation of
| a formula into computer hardware. It was corrected by adding several dozen
| transistors to the chip.

the error arose in transcription at some stage. Is that right? Perhaps I
read too much into journalese. But if so, it may be that formal
verification wouldn't have helped unless it was very tightly coupled with
the production process.

Nevertheless, I believe floating point hardware is a particularly good
target for formal verification by theorem proving. Providing the necessary
mathematical infrastructure for such verification efforts was the
motivation for all my work on formalizing analysis in HOL (though it has
since become an interest in itself).

John.
