So, what about the next 50 years?

The gains a small company could reasonably make [black line, a 75-percent-per-year, noncompounding rate] can’t compete with Moore’s Law [red line] if the doubling time is just18 months [left]. But a doubling time of 36 months [right] provides ample opportunity [yellow]. Credit: Andrew Huang, IEEE Spectrum

Yesterday I shared some considerations on the occasion of the 50th anniversary of the Moore's law. Today I would like to point your attention to a nice article that has been published on Spectrum on a future without the sort of evolution that has been epitomised by the Moore's law.

At first glance one might feel lost. We have been living our life under a perception of time clicking at the pace of Moore's law. Now it is starting to slow down and it might stand still.  The "might" refers to the fact that the continuous increase in performance au pair with a continuous decrease in price seems to have reach the end of the line. 

This is basically true for silicon. We do not know if new paradigm based on since layers materials, like graphene or molybdenum bisulphide, or spintronics, will be able to set up a new era of Moore's like evolution. As Richard Feynman said "There is plenty of room at the bottom" and we are still far from reaching the bottom although economics seems to prevent most of the path leading to the bottom.

So, for the time being let's assume that indeed in the coming years the Moore's law will no longer apply (in its two aspects performance/economics).

The article I read claims that such a slow down would change the rules of the game and would actually bring to the fore new players and even boost the economy as a whole. This seems an unbelievable claim so let's take a look at it.

The Moore's law did change the rules of the game! Long time ago I remember that when you bought a radio you would get the electrical scheme of the circuits. That was used by the repair man to fix a faulty vacuum tube or transistor (I did some repair job myself). It was also used by small companies to upgrade the product. This is al in the past. It is now many years since I saw an electrical scheme included in the box of the product. 

The reason is that the evolution of electronics, of chips, made any improvement in the circuit moot. A new version of a chip would come so fast that there would be no time to sell an improved version of the product based on the "old" circuit. It made more sense to wait for the new generation of chips.

This effectively killed any idea of "open hardware" and pushed small companies at the edges of the product to provide add ons.

The situation changes with the slowing down and demise of the Moore's law. As shown in the figure, even a slow down from 18 months to 36 months would be enough to open an "economic" opportunity windows for small companies and longer evolution times would even push the whole economic system to support "open hardware".

Interesting one of the closing remark of the article. In the future, may be in the third decade of this century, you might end up buying a personal computer (assuming you will buy something that will still be called a PC, which I doubt) and your grandchildren will inherit it, as we have inherited a precious watch from our grandfather. An amazing thought, isn't it?

Author - Roberto Saracco

© 2010-2020 EIT Digital IVZW. All rights reserved. Legal notice. Privacy Policy.