[Lawrence Phillia]

How much computing power does it take to model the brain, neuron by neuron, using conventional methods ?

Ccortex appears to be telling us something.

http://www.ad.com/images/press_photos/ccortex.jpg

[JamesL]

I believe the answer is a lot less then you would think. The problem however is that you would need a completely different kind of computer.

Its not a matter of how many switches, but the way in which the switches are arranged.

Ultimately things will have to be arranged in parallel. Given computer processes are a lot faster than neurotransmitters (say a thousand times) You only need to make the computer 1/1000 as parallel as the brain...... But you still have to make it parallel....

The question is how....

[Lawrence Phillia]

Phew! finally a response, that post's bin sitting there for about 2 years! thnx James.

[JamesL]

Interesting, can you tell us more about your work....

[Lawrence Phillia]

Short of spilling the beens ,sorta speak, I'd love to post one of my fledgling layouts and launch into an indepth explanation but I highly doubt it'd be intelligable. Perhaps some definitions would be much better.

http://www.cs.washington.edu/homes/diorio//Talks/InvitedTalks/Telluride99/index.htm

[DavidOlmsted]

I will know more in a few months as I really start putting my own software through some tests.
_________________
Click to go to my site at neurocomputing.org

[Lawrence Phillia]

Renee ( and others ), please post your questions in the forum.

Excuse my very brief intro to my research interests/work , a thorough discription would prolly require a manuscript, but i'll try ( at your request) to atleast give the essence of what im working on and I'll assume the reader has a knowledge of electronic circuitry.

First, im not designing/building a "microprocessor" in the Von Neuman sense of the word. It consists mainly of memory and a analog CPG ( central pattern generator) totally in hardware. Namely a bank of DS1250's which is a 4096K non-volitile SRAM IC, a MAX516 Quad comparator IC, along with some bilateral switches( Max 333,s),and a few LM198 sample and hold IC's which form the bulk of an integrate and fire adaptive CPG.

Second, the ADC/DAC combo ( analog-to-digital/digital-to-analog converter) was used as a means of "capturing" an analog value in digital form(ADC) and transmitting it via an address bus to onboard memory. During this same write cycle ( I use the term loosely since there is no clock ) if an analog update "collateral" is on the data bus, it is written into memory otherwise it is left un-disturbed . After training ,the circuit spends most of its time in a read cycle, i.e. digitized analog firing thresholds are continuosly sent ( via a DAC) to the CPG. IOW stored in memory are firing patterns, no real digital computation or processing as taken place.

finally, analog processing is an old idea which was displaced by more accurate and flexable ( tho slower and more computationally complex) digital processing. My research revolves around an hybred the two domains which i beleive is closer to real biological processing.

Hope this was an acceptable introduction.

[mikehahn]

I would like to create a simplified computer model of the neocortex, which I call a Toody Machine. Toody inhabits a 2D, computer-generated world, can see an array of pixels, can pan up/down/left/right, and can click on any of the pixels she sees. She can also hear/speak one ASCII character at a time.

I am a newbie to the field of modeling the brain in software, so my Toody Machine idea may or may not be wildly off the mark. I'd like to know what others think of it. To learn more about Toody, please visit:

http://www.treenimation.net/toody.html

Mike Hahn

[Intrepid Traveller]

Have you gents come across this study? http://www.damninteresting.com/?p=870

The article gives reference to the main research site.
I found it particularly interesting as the logic was not portable between like processors.

[Neil Hawkes]

Hi gents.

I have been doing some thinking in this area for a while - it may be of interest - as a way of looking at what can be made to happen with a Toody machine or an emergent system in an FPGA or in perception mechanisms more generally.

See http://positonicspage.blogspot.com/ for interest.

I would be keen to have your thoughts - it becomes necessary to seek like-minds at some point and I think that time is now for me.

Neil Hawkes