[RC5] new release candidate available.

Christophe Roguet christophe.roguet at free.fr
Tue Nov 26 00:24:10 EST 2002


Le lundi 25 novembre 2002, à 07:12 , Jason Hartzell a écrit :

> Yesterday's count is what's available. Come back tomorrow. Do you think
> 18,000 participants were shut off over the weekend? Even assuming 1/3 
> of the
> participants did not return stat units yesterday still gives you less 
> than
> 10,000 active participants. So that doesn't change or discredit 
> anything I
> said.
>
> Thanks.

I don't think comparing the number of daily participants to each project 
is a sound way of evaluating the actual participation in these projects

instead, let's compare the daily average computing power spent on each 
project

according to the plots on the stats page : 
http://n0cgi.distributed.net/statistics/ :

- at the end of RC5 64, the cracking rate was 200 Gkeys / s
- OGR is currently running at at least  300 Gnodes / s (and I guess it's 
more like 500 Gnodes/s but the plot is not precise enough to say)


now take a P III / 1Mhz as the reference machine

the calculator at this page : 
http://n0cgi.distributed.net/speed/graph.cgi  gives the client speed for 
both projects :

- such a machine can compute 3e+06 RC5 64 keys / s
- and it can compute 7e+06 OGR nodes / s


  so that gives :

- at the end of RC5 64, the cracking rate was equivalent to 66,000 
PIII / 1 MHz
- and currently, OGR is running at at least a rate equivalent to 43,000 
PIII / 1 MHz

that's only 1/3 below ... and again, this is with a low hypothesis of 
300 Gnodes a day ... the hypothesis of 500 Gnodes / s yields a power 
equivalent to 71,000 PIII / 1MHz ....


so, as Drew Scott Daniels  has pointed out before, the difference in the 
daily number of participants between RC5 and OGR is merely caused by a 
difference in the "granularity" of  the distributed work, and NOT by any 
supposed "boredom" of  the latter project  !

... maybe the stats page should be modified to account for all the work 
received over a complete (moving) week to better reflect the actual 
number of people working on OGR .....


this being said, when I take a look at this page :

http://n0cgi.distributed.net/statistics/ogr/index.html

I just can say that I am very eager to know the results of the work of 
the "dnet staff member who is not
involved in preparing the RC5-72 rollout" that Bruce Willson promised 
would be "posted to the web for all to see" in a message dated five 
weeks ago :
  http://lists.distributed.net/hypermail/rc5.Oct2002/0139.html


(and I'm certainly not whining .... just being curious !)


--
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