[cwjcc!hal!nic.MR.NET!tank!tartarus.uchicago.edu!kris@ohio-state.arpa: Benchmarks on LISP machines vs. work stations.]

[Mabry Tyson <TYSON@WARBUCKS.AI.SRI.COM>]

The note below is from AILIST.  I thought some of the people on the
Symbolics Lisp Users Group mailing list might be in a good position to
respond.  (I suggest responses be sent to kris@tartarus.uchicago.edu --
PLEASE do not just REPLY to this message as he won't get it.  I don't
know if this will spark discussion on AILIST but if so, I hope he will
summarize (to AILIST) messages sent only to him if they differ in
content than those on AILIST.  I don't see a real need for flooding both
SLUG and AILIST (and maybe the TI Explorer list, etc) with everyone's
personal opinions.)

I obviously have my own opinions (well, I *do* have a Symbolics on my
desk) but I'll avoid espousing them here.

Mabry Tyson


    Date: Fri, 9 Dec 88 13:10 PST
    From: cwjcc!hal!nic.MR.NET!tank!tartarus.uchicago.edu!kris@ohio-state.arpa
    Subject: Benchmarks on LISP machines vs. work stations.
    To: AIList@AI.AI.MIT.EDU

    I am looking at a choice between three basic alternatives:

	Running Common Lisp on Apollo 3500s ---
	TI Explorers and/or Micro-explorers ---
	Symbolics ---

    We're in an environment that is weighted heavily towards SUNs and
    Apollos, so there is a trade-off in terms of startup time and support
    for any movement towards dedicated LISP machines.  What I am looking
    for is any benchmark comparisons, discussion of the advantages of the
    Symbolics and Explorer envionrments as well as any horror/success
    stories that people have with both machines.

    Also, how are the Symbolics and TI environments for other functions
    besides lisp hacking?  How well do they integrate with SUNs? Apollos?

    Thanks in advance,

	Kristian Hammond
	kris@tartarus.uchicago.edu