Welcome to the Intel® Embedded Community.
Interesting concept. This community is focused on embedded systems and I would have never thought that could include embedding a computer into furniture. Gives a whole new meaning to "desktop computer", I guess.
Obviously making a larger, less dense processor is exactly the opposite of what the industry has been doing forever. Although I suppose it could be argued that multi-core processors are a step in that direction, e.g. integrating multiple, slightly slower processors on a chip rather than cranking up the speed and density of a single core.
Let's see if anyone has some ideas to share with you.
Good Luck with your research project,
Here are some of the challenges that you could run into.
- It will be insanely expensive to make a processor that is 2' X 2'. Think of all the silicon that it would require. I won’t be surprised if it cost 100X the current cost .
- Usually power and performance are tied together. To get more performance you need more power (or change the architectural design).
certainly i am not disagreeing with you, i just want to clarify- you are saying the high cost of processors is because of the raw materials, and not the fabrication process? what i had been hoping was as the fabrication would be much simpler and easier (considering the scale) that it would at least partially cover the cost of the silicon.
If that is the case, then, would it be feasible to run multiple processors using the same system? obviously taking advantage of space to solve heat issues. Assuming software isn't an issue, of course.
While there are obviously other raw materials included, Silicon is the second most abundant element on earth so the cost of a processor is not in the materials but the very complex and expensive fab process. I am sure that C-Akella's comment about silicon cost was not about a truckload of beach sand but rather the final cost of a silicon wafer. While you may be able to reduce the cost for a simpler design, a single wafer yields hundreds of processors (die) among which the cost is distributed. So the economics don't get you where want to go.
Here's a good tutorial on the chip making process:
By the way, Intel's leading edge fabs produce 300mm (approx 12-inch) diameter wafers, much smaller than your 2-foot idea. The next generation will be 450mm due in a few years. Historically, it has taken about a decade for the industry to migrate to the next larger wafer size, e.g. 200mm in 1991, 300mm in 2001....
Creating a system of multiple separate processors is a widely proven usage model. So you could easily pack a bunch of processors onto a 2-foot Printed Circuit Board (PCB), or even spread them around on separate modules networked together. Of course you need to factor in slower interconnect speeds compared to on-chip. For that matter you can buy commercial products today with multiple processors on a board. How well you could bury them inside furniture is another story....