[RC5] Project Order^M^M^M^M^Value

Zypher zypher at jknust.com
Sat Nov 27 01:52:09 EST 1999


-------------------------------------------------
***Warning:
You brain might implode upon reading this!
Keep a tylenol handy
...you have been warned***
-------------------------------------------------


Ok...I'll 'show my work' (groan) so someone else could substitute their own
values :) Digits past the 3rd or 4th I don't care about...nor does any
chemistry professor. (groan...sig figs)

If you managed a C or better in Calculus or Trig or something along those
lines, you can probably follow this...or my sanity is beyond hope.

I'll figure $2000 each for winning rc5 or csc. Euros might be more than 1=US
dollar, but after all the damn conversion fees and trouble (/disgruntled
former traveler who used credit cards and the like in many countries...and
found out the hard way) its $2000.

RC5 = 256 times bigger than CSC...I think. Whatever, but the stuff below
should be right even if I forgot shorthand exponential calculation. (pulled
out my TI89 for the rest :)
RC5 = $2000/2^64 = 1.084^-16 $/key
CSC = $2000/2^56 =  2.776^-14 $/key
Those are negative exponents, thus VERY SMALL VALUES.

My 450 Celeron does (when *idle*, taken from
http://www.planetfortress.com/rc5/benchmarks.htm, and no the Quad Xeon is
NOT mine, as if)

1337 *kilo*keys/sec RC5 (1.337 Million keys/sec)
561 *kilo*keys/sec CSC (0.561 Million keys/sec)

Too bad if your chip isn't P6-core based...bench it yourself.

Using le numerals and ye TI89,

RC5 = $ 0.000000000145 per second ...or $1 every 218.7 years :/
CSC = $ 0.0000000156 per second ...$1 every 2.03 years...about what I made a
couple years back ;)

CSC = approx. 108 times the 'revenue' of RC5...which matches my 256 times
smaller, 2.38 times slower (1337/561) figures...108 * 2.38 + ~257.

Of course only the winner gets the big bucks, but this is how much my
celeron 'contributes' to the prize pot, assuming 100% keyspace to hit the
key. Multiply by 2 (100%/50%) for 50% keyspace (average) and multiply by
[100%/(%age of keyspace completed when we find the key)] to find the
'actual' contribution rate...of course no one knows that %age yet :)

Ex: if it takes 22% to hit the key, then multiply 'contribution rate' by
(100/22)



(BTW, with dcypher, its the whole $10k ;) cough cough I guess I'm still
'loyal' to dnet though)

Oh wait...I'm on a team. Even though I'm the team coordinator, half goes to
winner and half spread among rest of team...at least thats how I'll do it if
we win ;) LOL

> Here is a little math puzzle for you:
>
> Estimate the value of processing each CSC and RC5 block.
> Measure the CSC and RC5 processing rates of your CPU.
> Calculate the ratio of CSC vs. RC5 blocks for your CPU that generates the
> highest value.
>
> -- Dan Oetting <oetting at ghtmail.cr.usgs.gov>

-------------------------------------------------
rc5 at planetfortress.com
Halflife/TF RC5
www.planetfortress.com/rc5
-------------------------------------------------


P.S.
Big deal if my typing is dyslexic...
So use me.

--
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