Re: BLUE GENE and confusion about flops and tops and whistles

From: Russell Wallace (russell.wallace@gmail.com)
Date: Fri Mar 31 2006 - 20:08:10 MST


On 4/1/06, turin <turin@hell.com> wrote:
>
> Excuse me, let me be more clear. I am not looking for a lesson in
> mathematics. I am in fact asking a question for SL4. I want to know why
> Russell Wallace said the CNS does 1-15 exaflops. I want to know his source
> or reasoning behind the claim. Compared to everything else I have read it
> seems very liberal.
>

Asking for the reasoning behind my claim is appropriate. 10^11 neurons, 10^4
connections per neuron = 10^15 connections, 10^3 Hz (commonly cited 200 Hz,
but it will turn out somewhat higher once you take in all the nuances, it
always does) = 10^18 ops, basic op is a multiply-add, and that doesn't take
into account the fact that it's actually more complex than that, the
connection modification (learning) dynamics (we don't really know how those
work yet), or all the other stuff we don't know about yet; 10^19 flops seems
reasonably conservative.



This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:00:56 MDT