h a l f b a k e r yIt might be better to just get another gerbil.
add, search, annotate, link, view, overview, recent, by name, random
news, help, about, links, report a problem
browse anonymously,
or get an account
and write.
register,
|
|
|
This idea is to develop a method for developing an addressing system and embedding logic chips along the length of the connecting wires and routers (the physical parts of the internet).
This might also include memory.
These would be made publicly available for massively parallel processing
activities and for changing code as it travelled along the internet.
The thing that might not be so obvious here is that you could choose a chip technology that is intrinsically cheap to manufacture, and let the fact that it's massively parallel make up for any performance compromises.
seti @ cell
http://www.halfbake...idea/SETI_20_40Cell [theircompetitor, Oct 05 2004, last modified Oct 17 2004]
[link]
|
|
So, you want to turn the wires that make up the internet into some sort of massive CPU? Instead of transistors, you use IP routers and switches? In this case, bigger ain't necessarily better. Count on clock cycles in Hz. |
|
|
Or are you thinking of distributed computing projects? |
|
|
What I'm thinking of is a concept that I've described rather poorly, but yes, at the heart of it is distributed computing. The idea is to make the information pathways themselves capable of a certain amount of (probably remedial) processing. Essentially embed a massively parallel supercomputer within the infrastructure. (I'll attempt to update the post with a better explaination tomorrow) |
|
|
There are so many devices and chips connected in some way already, the bigger challenge is to come up with OS that can handle it and problems that can be solved this way. |
|
|
I, too, am having a hard time figuring out what the job of these embedded chips would be. |
|
|
What is the scenario under which you see this being helpful? What value would the additional processing supply? "Changing code as it travelled along the internet" doesn't really strike me as being a good thing. |
|
|
There's a singularity point where processing power in the Net approaches and then passes human intelligence's processing power -- but I think we're going to get there without necessarily making every wire "intelligent" |
|
| |