Massive PDP (was: Lessons)

From: mAsterdam <mAsterdam_at_vrijdag.org>
Date: Wed, 14 Feb 2007 02:00:26 +0100
Message-ID: <45d25f17$0$337$e4fe514c_at_news.xs4all.nl>


Marshall wrote:

>>> What big questions remain unanswered in your mind?
>>   1. What are the biggest challenges to distributed optimization?

>
> An interest of mine is abstractions for writing
> highly parallelized/highly-distributed software.

http://www.scism.sbu.ac.uk/inmandw/tutorials/pdp/pdpintro.html

http://mitpress.mit.edu/catalog/item/default.asp?tid=7754&ttype=2

While the main living RM proponent Chris Date doesn't seem to care there is a promise in RM for neural networks, even though, as it is stated, is is a single-point-of-information model. My hunch is that appropriate hardware (no, simulations on single/few processors just don't cut it) needs to be accessible to garage-workers before we see how this works out. Received on Wed Feb 14 2007 - 02:00:26 CET

Original text of this message