computernewb

Distinguished
Aug 9, 2010
1,025
0
19,360
hi. i just came across this article claiming that moores law will end in about 10 years.
http://www.foxnews.com/scitech/2012/04/30/moores-law-ends-in-10-years-physicist-claims/

Michio Kaku is very intelligent and i have a lot of respect for him but im not sure how accurate his claims are. What are the opinions of the experts in the computer industry? I mean i can see that the bulldozer was a fail and the ivy bridge is a minor improvement to its predecessor but was it a 1 time fluke? or is Kaku's claims true?
 
Solution
I love Moore’s Law, or should we call it Moore’s principle. That every 18 months that the number of transistors on a processor will double. For years I have been year from one person after another that we can’t keep up with Moore’s Law. I am sure that sometime in the future are we are not going to be able to keep getting smaller but it isn’t for some time yet.


Christian Wood
Intel Enthusiast Team

Plasmid

Distinguished
Dec 5, 2011
148
0
18,690
I don't really have a source, but I read that intel engineers found that the limit was 16 nm ( 5nm gates) and that progress could not be achieved without quantum tunneling whatever that is.

I did some google search and ironically haswell (next gen intel cpu ) will be made on 22 nm process. Seems like intel is hitting a wall on how small they can get.
 

Stupido

Distinguished
Sep 17, 2008
342
0
18,810
I love Moore’s Law, or should we call it Moore’s principle. That every 18 months that the number of transistors on a processor will double. For years I have been year from one person after another that we can’t keep up with Moore’s Law. I am sure that sometime in the future are we are not going to be able to keep getting smaller but it isn’t for some time yet.


Christian Wood
Intel Enthusiast Team
 
Solution

loneninja

Distinguished


How is it ironic? They're following their tick tock strategy since 2006. Refine the current architecture on a new process node(Ivy is the new process node of Sandy), than release the next new architecture on that same process node, than repeat.

Core started on 65nm, was shrunk to 45nm. Nehalem was than released on 45nm, and shrunk down to 32nm for Westmere. Then Sandy came out on 32nm, and now has been shrunk to Ivy on the new process. Nothing abnormal happening here, Haswell should be 22nm.