At 09:33 AM 9/29/99 , Bruce Toback wrote:
>PS. Incidentally, while this thread is off-topic relative to HP 3000s,
>it's not off-topic relative to computing. SETI@Home is as much an
>experiment in dealing with very large computational tasks as it is an
>experiment in exobiology. Anyone who's entered the workforce more than
>about five years ago not only tends to reject solutions that have
>SETI@Home-sized computational requirements, but probably doesn't think of
>them in the first place. Yet there are probably business problems going
>begging right now that could easily be solved with large amounts of
>computational power.  Most of us need to switch to a mindset that says,
>"computer power and storage is free". It's blazing hard to do that when
>you've spent most of your career conserving cycles and bytes, yet that's
>now the wrong way to produce effective solutions to business (or
>engineering) problems.

What Bruce said.  :)

I haven't joined the SETI@Home project yet because I have my machines spare
CPU cycles working on Distributed.net projects and have been doing so since
late 1997.

Presently Distributed.net is working on an RSA sponsored challenge against
64bit RC5 encryption, but in the development and implementation pipeline
are enhancements to allow the Distributed.net clients to work on other
projects such as OGR (Optimal Golomb Rulers, a math thing with practical
uses in the real world :) and CSC (which is another encryption cracking
challenge sponsored this time by CS Communications and Systems).  Other
future potential projects include such things as a search for large
Mersenne primes).  D.net is the folks joined by the EFF to crack a 56bit
DES encrypted message in 22 hours early this year.  Also, all of the D.net
projects have a finite amount of work to do to make a milestone, whereas
SETI is a "we might someday if we're lucky stumble across something if the
something to stumble across exists in the first place" (which I think it
probably does, for whatever that's worth ;).

Distributed.net's client software runs on a variety of hardware and OS
platforms including all the major desktop machines and lots of multiuser
systems (including HP-UX on PA-RISC, but not on MPE yet though I was sent
the source before it was ported and optimized for PA-RISC).  Another
difference between Distributed.net and SETI@Home is that the D.net client
software has a much smaller footprint on memory, diskspace, and (I think)
network bandwidth.  But it's another mechanism in place harnessing the idle
computing power of many machines around the globe to solve problems that
are unfeasible with more monolithic architectures.

For (even) more info on Distributed.net see: http://www.distributed.net
And think about joining me on team # 1661, Libertarians for Privacy.  :)

Another project that may be more commercially viable in the long term is
called COSM.  See also: http://cosm.mithral.com
--
Jeff Woods
[log in to unmask]  [PGP key available here via finger]