Re: Do we really need virtual machines?

Paul Colin Gloster <Colin_Paul_Gloster@ACM.org>
21 Oct 2004 22:28:22 -0400

          From comp.compilers
Related articles
[9 earlier articles]
Re: Do we really need virtual machines? dot@dotat.at (Tony Finch) (2004-10-09)
Re: Do we really need virtual machines? Nicola.Musatti@ObjectWay.it (2004-10-09)
Re: Do we really need virtual machines? danwang74@gmail.com (Daniel C. Wang) (2004-10-09)
Re: Do we really need virtual machines? gah@ugcs.caltech.edu (glen herrmannsfeldt) (2004-10-09)
Re: Do we really need virtual machines? david.boyle@ed.tadpole.com (2004-10-12)
Re: Do we really need virtual machines? anton@mips.complang.tuwien.ac.at (2004-10-17)
Re: Do we really need virtual machines? Colin_Paul_Gloster@ACM.org (Paul Colin Gloster) (2004-10-21)
Re: Do we really need virtual machines? postmaster@paul.washington.dc.us (Paul Robinson) (2004-12-17)
Re: Do we really need virtual machines? gah@ugcs.caltech.edu (glen herrmannsfeldt) (2004-12-19)
Re: Do we really need virtual machines? tobias@berg.dichter.de (Tobias Bergmann) (2004-12-22)
Re: Do we really need virtual machines? vbdis@aol.com (2004-12-23)
| List of all articles for this month |
From: Paul Colin Gloster <Colin_Paul_Gloster@ACM.org>
Newsgroups: comp.compilers
Date: 21 Oct 2004 22:28:22 -0400
Organization: Compilers Central
References: 04-10-042
Keywords: VM
Posted-Date: 21 Oct 2004 22:28:22 EDT
Content-Disposition: inline

Scott A. Moore said:


"In addition, despite the fact that most people were reminded of P-code
when Java came out, Sun attempted to put the JVM on silicon and
achieved EXACTLY the same results Western Digital did when they tried
to put the P-machine on silicon, namely a processor that could not run
bytecode as fast as a general purpose machine. Turns out all that
register manipulation and multiple modes makes a difference after all.


The tired saying that applies is: The first thing you learn from
history is that nobody learns from history."


Less famous examples have also occurred. For example, SAAB created a
processor intended to be optimized for Ada but a document (written by
a rival) claimed better performance. It is located at
pub/ws/wsd/erc32/doc/ada_bench4.pdf on FTP.ESTEC.ESA.NL (by FTP
obviously).


This kind of attempt at optimizing had been discouraged in "Computer
Architecture: A Quantitative Approach", 2nd edition, by John L
Hennessey and David A Patterson, Morgan Kaufmanm Publishers Inc, 1990,
1-55880-069-8.

Post a followup to this message

Return to the comp.compilers page.
Search the comp.compilers archives again.