December 2nd, 2003, 03:24 AM
So long Moore's Law
I did a quick search on Moore's Law and didn't find anything so I thought I would post this link I found on SlashDot .
End draws nearer for Moore's Law
Basically for those of you who havn't heard of Moore's Law, the law states (roughly I'm no expert) that the amount of transistors that will be able to fit on a chip will double every two years. This explains the reason for computers becoming so much more powerfull and cheaper in recent years. Think of it, doubling every two years that is an expadential growth.
2...4....8....16.....32.....64......128 it grows fast is what I'm getting at!
The article talks about how in the somewhat near future, researchers beleive that they will hit a wall when trying to reduce the size of the transisters. The problem lies in the size of the gate (measured in nanometers) will become so small that the possibility of the electron in the chip passing through the gate even when the gate is closed will become to great. Another major problem designers face is energy consumption and heat given off by the tranisitor packed processors. All in all an interesting read.
December 3rd, 2003, 08:47 PM
I believe that one of the reasons for the discontinuation of this is because of the advent of the 64 bit processor. As many of us know by now, the 64 bit CPU will run over four billion times faster, or at 2^32*2^32 bit processors. I believe that because of this, it is no longer needed to double the number of transistors every two years. However, I believe that Moore's law may kick in again a year or two after 64 bit processors become the standard, and are used up to their full potential.
I may, of course, be wrong about this. Anyone else have any input?
December 3rd, 2003, 11:41 PM
Actually I think the answer is one of semantics (meanings of words)
It should be Moore's "theory" not "law"
In physics a "law" is immutable, cast in stone, never fails. A "theory", on the other hand will explain events in a certain or particular set of circumstances.
I think that we are just getting a little sloppy in our use of scientific English?
As rightly observed earlier, the environment has changed (64-bit) so the theory does not hold good in its original form?
Just my thoughts
December 3rd, 2003, 11:59 PM
people have been banging on about Moores 'law' failing for as long as I can remember - certainly since the late 80's. This page might interest some of you http://www.arstechnica.com/wankerdes.../limits-1.html
Quis Custodiet Ipsos Custodes