Physicist predicts Moore's Law will collapse in about 10 years

[Yazbuh]

---

Does this means in decades time, it will be the end of Silicon powered computing? Silicon chips reaching its limits? what you think guys?!

Theoretical physicist Michio Kaku has posted a video discussing the collapse of Moore’s Law in about 10 years or so. Physicists have been predicting the end of Moore’s Law for quite some time but Kaku’s reasoning and the slowing down of processing power that we already see today lends some credibility to his claims.
For those unfamiliar, Moore’s Law pertains to computer hardware, stating that the number of transistors that can be placed on an integrated circuit board can be doubled roughly every two years. You’ve probably alternately heard of an 18-month time frame tossed around. This modified cycle can be credited to Intel executive David House, not Moore.
Kaku says that in about 10 years, silicon power will be exhausted. Intel has already admitted that Moore’s Law is slowing down using silicon which is one of the reasons that we are seeing Tri-Gate transistors used in Ivy Bridge CPUs – an effort to try and extend the effective life of silicon.
The problem, Kaku says, is two-fold: heat and leakage. Today’s Intel processors have a layer that is almost down to 20 atoms across. When this layer is shrunk down to around five atoms across, “it’s all over.” At this point, the heat that is generated will be so intense that the chip will melt. The other concern is leakage, meaning we simply wouldn’t know where the electron is anymore.
So what’s in store for computing in the post-silicon era? Multiple proposals have been laid out, including but not limited to optical computers, protein computers, DNA computers, molecular computers and quantum computers.

Video


The latter two solutions seem to be the most commonly accepted “future computers” but both solutions still present enormous challenges that must be conquered. Molecular computers already exist but mass production and wiring up the tiny molecules remains an issue.
Quantum computing is even more finicky with the world record for a quantum computing calculation being: 3 x 5 = 15. As Kaku explains, it doesn’t sound very impressive until you realize it was proven using only five atoms.
Wrapping it all up, Kaku predicts that scientists will tweak Moore’s Law in the next 10 years to extend its life. After that, molecular computers will likely take over followed by quantum computers later in the 21st century.

Source

[lloyd_joy]

let's wait and see lang for processors... but, sa mga road maps sa semicon industry... on mass produced products, wla pa'y lain plano na materials in the manufacture of the "integrated circuits" gawas sa silicon...

silicon design and manufacturing is hard to understand... for me, personally, I don't buy the idea na exhausted na ang "silicon power" for the CPU wafers & die unless time comes na mao na gyud ni... its hard to predict when...

[edoy]

DNA computers and molecular computers was a concept more than 20 years ago pa, pero wala pajud na materialized. basin walay kapitalista mo gamble ani.

[Yazbuh]

I think exponential increase in in performance every two years is slowing down, maybe silicon powered computer in almost at its limits, mao siguro init kaayo temp sa ivy-bridge
Lets see what those engineers would com up with in a decades time

[Yazbuh]

[hawkshock]

Mao diay hehee failed ang IB! hapit na ma end ang silicon powered computing (Moore's Law)....

[renaire33]

Mao diay hehee failed ang IB! hapit na ma end ang silicon powered computing (Moore's Law)....

it not fail. just what kaku said, moores law is slowing down, as what we have seen in ivy bridge, temps are too high because of the chip size,silicon power is almost exhausted.

[lloyd_joy]

nope, as i'm in the semicon industry... dili ni mao ang reason... naa'y something sa design sa intel why init ang IB micro-architecture inig OC... OC raman pud... at stock speed, its temp performance is as expected. maybe, there is something in the silicon process or the circuit design itself... time will tell... ma solve ra ni nila... just give them time... they have the luxury of time since their main competitor, AMD don't have an offering that can compete head to head w/ them (Intel) as of this writing... this is a fact, right?

naa ko sa Quality Assurance and Reliability sa industry... I can empathize with Intel peeps... kun tinuod ni na problem sa design or process, ang QRA nila, sige'g pang latos ron sa process & design eng'g as well sa subcons para sa problem's resolution... hehehe... kana kun proven na problem gyud ni ha...

by the way, mao ni akoang buhat man pud gud... QRA are the "devils advocate" sa company...harharharhahar...

[Yazbuh]

Naa jud siguro sa design ila problem ani, since naa sila gipang add new blocks sa architecture sa Ivy, pero joke raman pud na ako about ivy hehehe
As they say in decades time paman silicon still has long way to go

[PagkahomotSAimongUTOToi]

It's one step every two years, and each step is 70.7% size of previous (square root of ½), so roughly:
22 nm approximately now
16 nm in two years
11 nm in four
8 nm in 6
6 nm in 8
4 nm in 10
3 nm in 12
2 nm in 14
1.5 nm in 16
1 nm in 18
800 pm in 20
500 pm in 22
350 pm in 24
250 pm in 26
and so on...

but..
The physicist is right, its inevitable. d mn 10yrs from now pero im sure it will come (the end of shrinking), although layo pa or dugay pa na mahitabo. I think that there is no current technology that can manufacture beyond nanometer (correct me if im wrong). IB is not a failed processor/architechture(SB refinement), its actually a more faster version of SB.

reference:
Semiconductor device fabrication - Wikipedia, the free encyclopedia