Molycircs coming soon?

pushmar wrote:http://www.gizmag.com/single-molecule-transistor/38476/

Cool post. Thanks.

My question is how power efficient will the tech be? Of course, I'm hoping for tablets with 3+ GHz cpus featuring 12 hr battery capacity on highest screen brightness playing videos or the heaviest cpu load. And... small and very light, running cool as the lakefront morning breeze...

Somewhere over the horizon, somehow. With some tech. Small just doesn't mean super efficient. Drats.

Oh my, I wonder how Apple will take a bite outta that?

I'd just be happy to see a Mac in tablet form.

Think of the built in redunctcy. Each chip could have a thousand parallel routes of operation. That means such a device would be very very reliable indeed with a very long life.

Combine that with a thousand or more chips in the gadget employeed and it might mean the electronical gadget would be basicely bullet proof. Non destroyable. Wil not quit.

Take that same theme with starship design and contruction and you would have a war ship that literally could take a licking and keep on ticking. Sorry about speeling. HB

HB of CJ wrote:Think of the built in redunctcy. Each chip could have a thousand parallel routes of operation. That means such a device would be very very reliable indeed with a very long life.

Combine that with a thousand or more chips in the gadget employeed and it might mean the electronical gadget would be basicely bullet proof. Non destroyable. Wil not quit.

Take that same theme with starship design and contruction and you would have a war ship that literally could take a licking and keep on ticking. Sorry about speeling. HB

Computer chips already count their number of transistors in the low billions, do you really expect every one of those to be doubled up a thousand times?

No matter how small, every single transistor costs something to create, and the materials used are also starting to become rather on the pricey side. And making a chip 1000 times larger than needed would increase the manufacturing and materiel cost by at least a 100 times, probably much more.

And chips using the linked transistor would still only be something like 1/5 to 1/10th the size of chips made today(Broadwell is made with 14nm process).


More troublesome is that the article doesn´t really provide solutions for the problems shrinking is causing.

And it´s just a few weeks since Intel stated that for the first time since 2007, they will not be able to proceed with their normal "schedule", which would have meant launching Cannonlake chips made at 10nm process in 2016, instead there will be another 14nm chip, Kaby Lake in 2016.

This also comes after their complete failure at launching Broadwell on schedule, instead pimping up the 22nm Haswell refresh chips for 2014 and then launching Broadwell only a few months before its replacement Skylake takes over on August 5th.

In short, Intel is having BIG problems even now to make the chips reliably in massproduction.

Meanwhile, AMD is still using 28nm process and will shift to 14nm 2016 or 2017.

There is even some degree of uncertainty whether 10nm will happen at all.

Also meaning that something as small as the gate in the article may not be useful at all.
No matter how well it works in isolation, it may not be possible to make it work together with a few billion others on a chip a sqcm or two big.

Controlling electron migration and leakage is going to be outrageously difficult.

More likely we will see one form or another of 3D stacked chips long before we see transistors below 3-4nm.

A few people around work are more worked up about this potentially panning out:

http://www.princeton.edu/main/news/arch ... topstories

An international team led by Princeton University scientists has discovered an elusive massless particle theorized 85 years ago. The particle could give rise to faster and more efficient electronics because of its unusual ability to behave as matter and antimatter inside a crystal, according to new research.

The researchers report in the journal Science July 16 the first observation of Weyl fermions, which, if applied to next-generation electronics, could allow for a nearly free and efficient flow of electricity in electronics, and thus greater power, especially for computers, the researchers suggest.

Proposed by the mathematician and physicist Hermann Weyl in 1929, Weyl fermions have been long sought by scientists because they have been regarded as possible building blocks of other subatomic particles, and are even more basic than the ubiquitous, negative-charge carrying electron (when electrons are moving inside a crystal). Their basic nature means that Weyl fermions could provide a much more stable and efficient transport of particles than electrons, which are the principle particle behind modern electronics. Unlike electrons, Weyl fermions are massless and possess a high degree of mobility; the particle's spin is both in the same direction as its motion — which is known as being right-handed — and in the opposite direction in which it moves, or left-handed.

...

(ctd. at link...)