IBM researchers have made a dramatic new breakthrough, creating a new model for
the transistor that could one day lead to less power-hungry computer chips that
mimic the way the human brain operates.
Transistors have been at the heart of tech-heavy society for
decades – they can be likened to the ‘building blocks’ of the electronic
circuits that power all of the gadgets we take for granted nowadays, be it
smartphones, tablets, PCs or televisions. But despite all of the technological
advances made in recent years, with ever smaller but more and more powerful
devices emerging each year, the transistor itself has seen little change, until
now.
The breakthrough, which is described in the journal Science and reported by The New York Times, involves a new coating for the transistor that allows it to read ionic signals, rather than standard electronic ones. Older transistors have traditionally relied on electrical currents to switch materials between conducting and insulating states, but the new method uses what IBM describes as “ionic currents”, or mobile charged atoms, as a switching mechanism. Essentially, circuit boards built using the technique would contain ionic liquids, a kind of mixture where half of the molecules are positively charged, and half are negatively charged, similar in many ways to the electrical fluids that slosh around in the human brain.
The breakthrough, which is described in the journal Science and reported by The New York Times, involves a new coating for the transistor that allows it to read ionic signals, rather than standard electronic ones. Older transistors have traditionally relied on electrical currents to switch materials between conducting and insulating states, but the new method uses what IBM describes as “ionic currents”, or mobile charged atoms, as a switching mechanism. Essentially, circuit boards built using the technique would contain ionic liquids, a kind of mixture where half of the molecules are positively charged, and half are negatively charged, similar in many ways to the electrical fluids that slosh around in the human brain.
Admittedly it all sounds a bit technical, but IBM says that
it’s a very big deal as the development should allow chipmakers to increase the
number of transistors it can load onto a single chip, vastly increasing their
processing power.
Big Blue’s research team says that they haven’t actually
made a chip using the new transistor, but it is able to demonstrate how a rough
circuit would work. Even so, the company says that the new technology should be
commercially viable, and expects it to leave the lab in the next five to seven
years.
Do We Need a New Transistor?
We sure do, if we want to keep on
innovating and building smarter new devices. One of the problems faced by the
chip industry today is keeping up with Moore’s Law,
which dictates that the number of transistors on a chip should be doubled every
18 months to two years in order to foster new innovations. Up until now, chip
makers like Intel have come up with the goods just by stuffing more transistors
onto their chips each time, but doubling up is becoming much harder and its
uncertain how long they can keep doing this.
Some might argue that we don’t need to stick Moore’s Law,
but the fact is if we want to see technology progress then costs have to be kept
down – if electronics become more expensive, then companies like Facebook and
Google will find it harder to innovate.
IBM’s new technique, if it really does prove viable, should
help to deal with one of the biggest problems presented by cramming smaller
transistors on a chip. Electrons simply aren’t that efficient, they make lots of
heat and noise, and so they drain lots of power. Hence, the more electrons on a
chip, the more expensive that chip becomes. By using ionic fluids instead of
electrons, IBM says that it can do away with much of this ‘leakage’, which would
allow chipmakers to keep on doubling up without worrying about the
costs.
By:
Aucun commentaire:
Enregistrer un commentaire