[rc5] PII figures (sorta)

Paul Leskinen paul at csfi.com
Mon Jul 7 15:20:15 EDT 1997


Interesting to note that if you "scale down" the figures to simulate a 200
MHz Pentium II, they're about the same as the Pentium Pro:

v2.0p6 Win32 PPro: 455979
v2.0p6 Win32 PII:  444556 (+3% fudge factor = 457892)

At best, the PII has only a very modest speed improvement: 0.4% faster than
equivalent PPro (which is less than the deviation of the test).  Even with
the 611K figure, it's only 1% faster than the PPro.  But still an _awful_
lot of keys per second.  Should finish a block every 7.3 minutes!  That's
over 2 billion keys per hour.
 
Verrrry interesting. Hey Intel, I'll take a PPro 266, please, for less
cash.  Oh, you don't make those? Hmmph.

----------
From: Bob.Rich <Bob.Rich at Enterprise.globalec.com>
To: rc5 at llamas.net
Subject: [rc5] PII figures (sorta)
Date: Monday, July 07, 1997 1:43 PM

The 'sorta' is because I had Tim Charron's 'hidden' v1 client running on
the box at the same time (doing real work), but on the PPro that was
previously posted, this resulted in a very consistent 3% slowdown, so we
can just fudge.

Machine

Dell Dimension XPS H266 Pentium II
64MB RAM
6GB IDE Disk

Only running the Win32 clients, no Linux (yet!)  Five runs, high/low cut,
again very consistent numbers (even had one that was _exactly_ the same kps
rate)

Client			Real			Fudged (+3%)

v2.0p5			587471			605095
v2.0p6			591259			608996
v2.001p5		587279			604897
v2.001p6		590947			608675

My 'fudged' numbers must be off a bit, because i KNOW i've seen 611000kps+,
but this will at least give ballpark figures.

I wonder if 266's can be overclocked. ;)


- Knock-knock
o Who's there?
- Impatient cow
o Impa...
- Moo!

:)




----
To unsubscribe, send email to majordomo at llamas.net with 'unsubscribe rc5'
in the body.

----
To unsubscribe, send email to majordomo at llamas.net with 'unsubscribe rc5' in the body.



More information about the rc5 mailing list