Path: utzoo!utgpu!water!watmath!clyde!att!osu-cis!tut.cis.ohio-state.edu!mailrus!uflorida!gatech!hubcap!fpst
From: fpst@hubcap.clemson.edu (Steve Stevenson-Moderator)
Newsgroups: comp.parallel
Subject: Memory access conflicts
Message-ID: <2887@hubcap.UUCP>
Date: 29 Aug 88 12:26:41 GMT
Sender: fpst@hubcap.UUCP
Lines: 19
Approved: parallel@hubcap.clemson.edu

> There are situations in which it is likely that processors operating in
> parallel may have to address the same item in memory at the same time.
> What are the consequences in program time for conflicts of this type?
> What hardware features are there to possibly circumvent this problem if
> it is serious?

This is a very old and serious problem.  It has occupied an lot of
pages in the performance literature.  Hardware folks became aware of it
early on.

Given your ``residence'' you might try Coffman and Denning "Operating
System Theory" --- old but has all the queuing stuff in  it.  

There are probably newer treatments, but the problem's the same.

Steve Stevenson                            fpst@hubcap.clemson.edu
(aka D. E. Stevenson),                     fpst@prism.clemson.csnet
Department of Computer Science,            comp.parallel
Clemson University, Clemson, SC 29634-1906 (803)656-5880.mabell