Subject: Re: PC164 memory bus speeed (was: pciide performance on alpha)
To: None <port-alpha@netbsd.org>
From: None <seebs@plethora.net>
List: port-alpha
Date: 11/02/1999 14:16:00
In message <199911021939.OAA10085@ghost.whirlpool.com>, Andrew Gillham writes:
>Does anyone have a list of PC164 optimizations that should be turned on?
I'd like to see these too.
>I just got a 500Mhz PC164, and while being fairly impressed with it's
>speed,(ok I also have a 166Mhz Multia) I was surprised that my 433Mhz
>Celeron generally beat it.
Don't be. The Pentium Pro, PII, and Celeron all have an architectural
advantage over the 21164. Intel came up with this wacky idea called "out of
order execution", where the processor reorders instructions as they come to
make them more efficient. (This has to be done on the chip, not in the
compiler, because "more efficient" can depend on what happened on the other
side of a context switch.) The 21264 has this feature too, and other related
hacks, but the '164 is fairly old.
>Tuning tips, performance numbers, or CISC .vs. RISC flames accepted. :)
See above. That said, I'm quite confident that improvements could be made,
and I'd love to see advice on what they'd be. Thanks to some heroic efforts
by a couple of NetBSD people, I've finally got my Alpha "debugged enough" that
I'm starting to use it as a production machine. :)
-s