[RC5] Die SETI die !
Matthew.Toseland at btinternet.com
Mon Jul 19 00:05:25 EDT 1999
OK OK I'm sorry for bringing it up yet again but...
Visit their stats Totals page and do the math... <groan> EIGHTEEN
TERAFLOPS (sorry for shouting, and it's just an order-of-magnitude thing,
assuming 200MFlop/CPU) ! I'm pulling in my other recently-disemboweled machine
as soon as I can get it working, and I'm sure this will make up the difference
:) We've been outgunned... start recruiting technophobic paranoid
schitzophrenic bosses who used to think it's a trojan.. :\
Total CPU time
263728038 hr 04 min 27.9 sec
Average CPU time
per work unit
33 hr 20 min 52.6 sec
Copyright 1991, 1992, 1993, 1994, 1997, 1998 Free Software Foundation, Inc.
This is free software with ABSOLUTELY NO WARRANTY.
For details type `warranty'.
Of course, on the other hand, this means that distributed computing is really
kicking arse, already beating IBM's 10-teraflop beastie scheduled for 200[1,2?]
- that's the only thing that really matters, as long as there is more
horsepower on dc than being used for virtual bomb tests :)
A lot of people in the "target audience", including me, would be much more
happy with full source... why is it so hard to publish source and double-check
0.1% or so of the blocks assigned? - one in every few hundred blocks raises a
false alarm, so get the keymaster to check that when a block has been sent in
with a false alarm flag, if a dupe of it comes in, if that has not got the
false alarm flag it will squeal about it; it will also occasionally
deliberately send out dupes to check this. It would only need to check the
status of those dupes, and on the first multiple-confirmation it could forget
the block as nobody is going to forge blockhoarding since dupes don't get
credited to your keyrate...
To unsubscribe, send 'unsubscribe rc5' to majordomo at lists.distributed.net
rc5-digest subscribers replace rc5 with rc5-digest
More information about the rc5