home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: comp.parallel
- Path: sparky!uunet!destroyer!gatech!hubcap!fpst
- From: kielmann@isa.informatik.th-darmstadt.de (Thilo Kielmann)
- Subject: Re: Glossary of Parallel Computing Terms
- Message-ID: <1992Sep15.150231.28876@news.th-darmstadt.de>
- Sender: news@news.th-darmstadt.de (The News System)
- Nntp-Posting-Host: sun24.isa.informatik.th-darmstadt.de
- Organization: TU Darmstadt
- References: <1992Sep14.122934.16072@hubcap.clemson.edu>
- Date: Tue, 15 Sep 1992 15:02:31 GMT
- Approved: parallel@hubcap.clemson.edu
- Lines: 32
-
- In article <1992Sep14.122934.16072@hubcap.clemson.edu>, Greg Wilson <gvw@epcc.edinburgh.ac.uk> writes:
- >
- > The second version of a glossary of parallel computing terms is now
- > available by anonymous ftp from:
- >
- > ftp.epcc.ed.ac.uk:pub/personal/gvw/glossary.ps.Z.uu
- >
-
- First of all: Thank you very much for this piece of work!
-
- Second: A question to your definition of the "parallel balance point".
-
- You define the time T taken to perform some calculation of size N on P
- processors as:
-
- T = fx(N,P) - fc(N,P)
-
- where the execution time function fx is a decreasing function of P,
- while the coordination time function fc is an increasing function of P.
-
- I worry about the subtraction of these 2 functions. Isn't it necessary
- to perform both execution _and_ coordination? So, why not add fx and fc?
-
- Is it just a typo, or am I wrong?
-
-
- Thilo Kielmann (kielmann@isa.informatik.th-darmstadt.de)
-
- Dept. of Computer Science (FB 20)
- University of Darmstadt
- Germany
-
-