No sector has experienced such rapid
development as the computer technology. This is well reflected in the Moore’s
Law. In essence, this law says that the number of transistors in a chip would
be doubled each 2 years about. This Law has been meeting exactly since it was
enunciated in 1965. To make a meaningful comparison, if this law was applied to
commercial aviation, a flight between New York and Paris would last less than 1
second and would cost 1 dollar cent (Wikipedia)!
But, In my opinion, new truly revolutions over
computational science will come before from new architectural and computing
paradigms than from maintaining the validity of Moore Law another 50
years.
The current manufacturing technology
chip's seems to be reaching new boundaries which could require reviewing the Moore Law.
The current level of chip’s integration is reaching the scale where quantum
physics effects should be considered. Specifically at less than 10 nm scale the
electrons are no longer confined in the ‘highways’ of the chips. This is known
as Tunnel Effect.
Scientists and leading companies are working to
Moore’s Law continues to meet. Quantic computation, information storage in DNA
(http://www.prensalibre.com/vida/ciencia/Cientificos-codifican-sonetos-Shakespeare-ADN_0_853714788.html) or optic computing, both still in
a very preliminary stage), could dramatically increase the computation capacity
as well as chip’s size.
I am recommending you watch this 2 minutes
video to understand intuitively what quantum computing is: http://youtu.be/rBdGPTLfskI
But, there is another interesting perspective
to explore. The architecture of current computers is very similar than first
ones. Almost all computers are based on the Von Neumann architecture since 1945
(http://es.wikipedia.org/wiki/Arquitectura_de_von_Neumann) and the fundaments of computer
processing are, essentially, the same ones as those days.
What do you think? Which will come first? Hardware, architectures, both?
No hay comentarios:
Publicar un comentario