[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [not about] gc-by-area and host uptimes

	This is precisely why Symbolics should be concentrating on software, and
	not hardware.  You are bound to lose if you are always behind the
	semiconductor giants.  Why try to beat them at a game you can't win?

	I can give you many examples of large systems that now operate
	significantly faster on a Sun4 than on a Lispm.  For example, with Lucid's
	new EGC, performance on GC heavy code has gone up very significantly.  I
	can compile *and* load several hundred thousand lines of code on a Sun4 in
	the *same* amount of time it takes to just load it on my 3645.  This is for
	a Lisp and EGC with no special hardware.  How do you explain that?

Alas, I don't have access to Lucid with EGC, but even without it, QPE
runs about three times faster on a Sun-4 ON SMALL EXAMPLES than on a
Symbolics.  The people who own Sun-4's around here configure them
inadequately (would you believe only 50MB of swap for everything??  I
sometimes run out with 200MB..), so I haven't been able to try large
examples on Sun-4's.

I have tried reasonably large problems on my RT (250MB of swap), and
it is neck and neck with my 3670.  The RT in question is a 125;
experiments with other programs indicate about a 25% performance
improvement with a 130, but I haven't tested it myself yet.

It wouldn't surprise me if going over the Symbolics system with an eye
towards streamlining and performance couldn't net at least a factor of
two speedup, given some of the kludgery I've seen in past sources.
That would help alot in keeping Ivory alive.