[rc5] Identical results, two CPUs
sean at ntr.net
Sat Jul 26 10:57:29 EDT 1997
> Among other machines, I'm running the client on a P200/MMX and a
> regular P200. A coworker says there's "no way" these two machines can
> be running at virtually identical speed, but they are - about 240Kk/s.
> Would anyone care to take a crack (ha ha) at explaining why they are
> the same so my coworker and I can end this silliness?
> Tom Wheeler
> tomw at intelligraphics.com
Sure; the client's efficiency and productivity are based MAINLY on the CPU
speed (I know there are some real cpu junkies that are gonna jump me for
this). They both are very similar chips in the way the basic portions of
the CPU runs and handles code; each chip will process roughly the same
number of cycles in a given amount of time. The MMX processors have the
special instruction set that was designed to let you see really fast
multimedia responses (hence the name - MMX MIGHT stand for Multi-Media
eXtensions, but intel swears it doesn't). These instructions don't do a
whole lot of good on the kind of basic number crunching we're doing.
That's why you'll see that a lot of programs that don't have any special
MMX aware code won't run any faster (except some of the newer MMX chips
have more cache on chip, but that again doesn't affect us much here). The
extra parts of the chip just aren't/can't be used for this kind of thing
(And I know you can maybe use them as programmable registers, etc etc etc,
but cut me some slack on this one, okay CPU gods? thanx).
I know you'll get a dozen of these responses, or more, but this is a kind
of basic approach to it that should satisfy the non-Uber Geek.
If they really wanna know more, check out http://www.x86.org or
sean at ntr.net
REALITY.SYS corrupted. Reboot universe? [Y n]
To unsubscribe, send email to majordomo at llamas.net with 'unsubscribe rc5' in the body.
More information about the rc5