[ExI] simulation

[hkhenson]

At 05:43 PM 11/29/2007, Bryan wrote:
>A distributed, open source brain simulation project could have many
>uses: scientific understanding, pushing for "open minds", a
>particularly long-term goal for the h+ programming community to work
>with, etc. Not everybody (read: you) have to support the concept, of
>course.

It's a worthy goal of course, but I think quite a way from being a 
project that could be done distributed, at least not on current 
machines and the current net if this bunch of researchers has the 
estimates of the machine capacity and bandwidth required correct.

To scope the problem, the area of the brain surface is about quarter 
of a square meter, the cortical columns are spaced about 0.03 mm in a 
hexagonal array.  Figure about a thousand of them to the sq mm. There 
are a million sq mm to a sq m. So if you had 250 H+ programmers, each 
of them would have to be managing a million cortical column simulations.

Please check my numbers.

If you have any suggestions about how to reduced the difficulty of 
the project, please post them.

Keith Henson