Linux On Steroids: DIY supercomputer software from Sandia

[Tim May]

On Wednesday, August 8, 2001, at 05:06 PM, Faustine wrote:

> On Wednesday, August 8, 2001, at 02:28 PM, Faustine wrote:
>
>> Lots of interesting possibilities for cryptographic applications, I'm
>> sure...
>> Massively Parallel Computational Research Laboratory
>> http://www.cs.sandia.gov/
>> Except when was the last time you heard of a Cypherpunks-interesting
>> cipher being broken with _any_ amount of computer crunching?
>
> Since when did people stop trying? The last time I heard a researcher 
> talk
> about trying to break a Cypherpunks-interesting cipher was last 
> Thursday.
> Hearsay and hot air? Probably; nothing that merits repeating. But it's
> hardly a dead issue.

And why not name who this researcher was, and why you think the cipher 
he was trying to break was Cypherpunks-interesting?

(Mere factors of 100 are not interesting, vis-a-vis good ciphers and 
large keys. Arguably, as we have discussed many times, not even 
"swimming pools full of Adleman DNA computers" make a difference.)

>> (The "challenges" broken by a couple of our own list members over the
>> past several years were all weak ciphers by modern standards, or had 
>> key
>> lengths way below even the recommended lengths of the day. Increasing
>> the key lengths by just several bits ups the work factor by a factor of
>> ten or so. Increasing it to recommended levels ups the work factor to
>> the level of "not all the computers that will ever be built in all of
>> the galaxies of the universe" will be able to brute-force a crack.)
>
> We've all heard that line before, but I still don't think it's too far-
> fetched to assume that anyone who does work in this area might 
> appreciate
> 50 megs of free software to create his own supercomputer.
>

Fatuous nonsense, Beowulf clusters have been out for several years. The 
hard part is getting 50 Pentiums, not the software.

As for "50 gigaflops," big whoop. A readily-available dual G4 machine is 
rated at about 8-10 gigaflops. (Or "FLOPS," as you wish.)

Still not interesting for cracking ciphers, in the real world.

>
> You never know what might come from putting that kind of computational
> power in the hands of people here. Create, break, do whatever you want.

You need to get up to speed, so to speak.


--Tim May