"patrik bagge" <patrik-b at online.no> wrote:
>[snip]
>>>The number of "neuron updates per second" for the biological brain is
>>listed as 10^14. So there is a 9 order of magnitude difference. Based on
>>the general speed increase over the last 4-5 years, that should be about 8
>>orders of magnitude by the present day.
>>>>If we assumed Moore's Law would continue to hold, this gives us an estimate
>>more optimistic than the "45 years for brain power on the desktop" estimate
>>posted earlier. By 2030, the heuristic gives a desktop PC with 10^12
>>neuron updates per second, 10^14 by 2040.
>>>>Note that these figures are for running some kind of (unstated) typical
>>neural net software on a typical desktop CPU. With some kind of
>>purpose-optimized system, the job could likely be done with significantly
>>less computing complexity.
>>>hi, i enjoyed reading your comparison between the human mind
>and a computer, but i have some 'amatourish' questions.
>>does your estimates take into account the processing power needed
>for the possible information flow in all these 1000 connections
>from every single neuron in the human mind?
Artificial neural networks usually have a significant number of connections
which have to be dealt with during updates as well. If for some reason
more connections had to be added, that would increase the required
processing time by some small constant multiplier.
>without knowing so, i assume that a human neuron works with
>analog signal levels, is this 'extra' A/D conversion taken into account?
Artificial neural networks are essentially analog as well. When run on a
PC, they use high-precision real numbers to simulate an analog signal.
