A new paper by Rodrick Wallace of the New York State Psychiatric
Institute suggests that multiple-core chip technology may lend itself to
implementations of recent mathematical models of the massively parallel,
high level mental processes in animals. The author notes that multi-core
computers present a programming problem that is becoming a nightmare as the
number of cores increases, with the potential to have machines with
thoursands or tens of thousands of cores. He proposes that one feasible
solution to the problem is to adopt the parallel computation model that
developed and was optimized by millions of years of evolution. One side
effect of doing this is that it may make machine cognition more
feasible. To get all the details, see Wallace's paper, Biologically inspired
distributed machine cognition: a new formal approach to hyperparallel
computataion" (PDF format).
Source: http://www.netchain.com
No comments:
Post a Comment