I was wondering if anyone has come across any credible information regarding a departure from silicon/transistors in the development of cpu's? Perhaps an ambitious research project somewhere in the world?
Most envision the addition of cells as biosensors for quick detection of compounds.
I don't think we will see a pure living chip that can out perform a solid state chip, due to size constraints. A neuron can be 10's of microns in size, and is axons maybe down to 100's of nm. Then there are the problems of providing the proper culture/living conditions.
Message edited by badgtx1969 on 04-06-2008 at 06:48:41 PM
---------------
Take what man makes and use it, But do not worship it, For it shall pass.
This is probably way over anyone's head but is really cool when looking at neurons and how they signal.
Neurons use a process called saltatory conduction to speed up signaling. There are insulating coatings called mylen sheaths, which are made up of mainly of fat, that run down the length of the axon. The sheaths create nodes (see the picture in this wiki link http://en.wikipedia.org/wiki/Schwann_cell) along the axon surface. The electrical signal jumps from node to node instead of traveling down the entire axon. This node to node behavior speeds up signaling tremendously opposed to traveling down the entire axon. Lack of saltatory conduction is actually the cause of MS. The mylen sheaths degrade and force signals to travel down the entire axon and considerably slows down signaling.
With the use of nanotechnology / nanobiology, it would be interesting to see if saltatory conduction could be mimicked and applied to current technology to speed up signaling. Imagine a wire that was a synthetic axon with mylen-like sheaths that forced electrons to travel faster via nodes rather than down the entire length of the wire. Faster signals and less heat is always a good thing. Someone should get on that.
I think the best option might be to copy some of the abilities of neurons, but as far as building a processor out of them it seems like it would be the wrong option.
Who knows? Maybe we will have Flux-capacitors in our PC's soon that run on old banana peels and beer. But we might need some of that nano tech in order to produce the 1.21 Jigawatts without mounting the PC on a Delorean and affixing a lighting rod to it.
By the way, where is Biff now anyways?
---------------
Core 2 Duo E6420 @ 3.2 ghz
SAPPHIRE HD2900XT (aka. The Fire Hazard 2)
2 Gigs Corsair DDRII 800 XMS2 DHX 4-4-4-12
Soundblaster X-Fi Platinum
sun microsystems is trying lasers. to badgtx1969: neurons are less than 20 nm. maybe you can use atomic or sub-atomic level precision for this and get almost the speed of light. -or if you use charged photons instead of transistors, the information would travel at the speed of light.
I can see it now, " My rig is sloooow" Answer: Have you fed your processor? Or, "Cant wait for my dual purpose water cooling/ processor feeding unit to come in from the egg"
hahaha. screw wiki. They have been successful in building logic gates with neurons, however, it's quite hard to get them to grow that way. So there will be scalability issues. Not to mention that nobody knows how to program them properly, or that neurons' axio-dendrite interface is way too analog to encode anything. Oh, and don't forget to install the latest microsoft patches with your handy dandy icepick.
Hmmm. Lets see. Bio computers. Fun times. Next thing you know there will be premium computer food that you feed for better performance. No more heat really. I would be using more of a Miracle Grow plant stick for my MS updates. Seems more appropriate. Then it dies when I go on vacation.