[RC5] iMac

Robert Zwaska rzwaska at nd.edu
Fri Dec 4 17:35:35 EST 1998

I've done some testing with G3 processors in general, and have come to a
few possible conclusions.

One, the G3 is new, and I don't believe that the client code has been
optimized for these processors.  This can have a very large difference. 
On  some 233 DT's I had trouble getting up over 650 kkeys/s, but this
was in real-world testing, so other processes were running.

The G3 might also not be in the best disposition for RC5 cracking. 
Note: with a 7300/180 (604e) I get about 600 kkeys/s.  In comparison, an
Ultrasparc 30 creater 3D (292 MHz I think) also gets about 600 kkeys/s. 
The Sparc is much faster than the Apple.  The problem is the built-in
instructions on the processors.

To eleaborate on this.  The G3 chip scores only slightly high MHz to MHz
than the 604e on many low-level tasks like integer and floating-point
benchmarks, but it scores significantly higher on many real-world
applications.  The G3 was designed with the idiosyncracies of the MacOS
in mind.  Thus it accelerates many tasks that do not need many
operations, but took a while to chug through the more general 60x chips.

Thus, I don't think you are going to see much of an improvement moving
from 60x chips to 65x chips.

Bob Zwaska				rzwaska at nd.edu
448 Zahm Hall				http://www.nd.edu/~rzwaska
Notre Dame, IN 46556			(219) 634-3601

To unsubscribe, send 'unsubscribe rc5' to majordomo at lists.distributed.net
rc5-digest subscribers replace rc5 with rc5-digest

More information about the rc5 mailing list