Researchers announce the discovery of an atomic electronic simulator

October 15, 2018, University of Alberta
Charge configurations of two closely-spaced DBs. Credit: arXiv:1709.10091 [cond-mat.mes-hall]

Targeting applications like neural networks for machine learning, a new discovery out of the University of Alberta and Quantum Silicon Inc. in Edmonton, Canada is paving the way for atomic ultra-efficient electronics, the need for which is increasingly critical in our data-driven society. The key to unlocking untold potential for the greenest electronics? Creating bespoke atomic patterns to in turn control electrons.

"Atoms are a bit like chairs that electrons sit on," said Robert Wolkow, physics professor and principal investigator on the project. "Much as we can affect conversations at a dinner party by controlling the grouping of chairs and assigned seating, controlling the placement of single atoms and electrons can affect conversations among electronics."

Wolkow explained that while atomic control over structures is not uncommon, making custom patterns to create new useful electronic devices has been beyond reach. Until now.

Though the tools of nanotechnology have permitted exacting control over atom placement on a surface for some time, two limitations have prevented practical electronic applications: the atoms would only remain in place at cryogenic temperature and could only readily be achieved on metal surfaces that were not technologically useful.

First proof of concept

Part atomic machine, part electronic circuit, Wolkow and his team have recently created a proof-of-concept device, overcoming the two major hurdles preventing this technology from being available to the masses. Both the robustness and the required electrical utility are now in hand. Additionally, the structures can be patterned on silicon surfaces, meaning scaling up the discovery is also easily achievable.

"This is the icing on a cake we've been cooking for about 20 years," said Wolkow. "We perfected silicon-atom patterning recently, then we got machine learning to take over, relieving long suffering scientists. Now, we have freed electrons to follow their nature—they can't leave the yard we created, but they can run around freely and play with the other electrons there. The positions the electrons arrive at, amazingly, are the results of useful computations."

Based on these results, construction has started on a scaled-up machine that simulates the workings of a neural network. Unlike normal neural networks embodied of transistors and directed by computer software, the atomic machine spontaneously displays the relative energetic stability of its bit patterns. Those in turn can be used to more rapidly and accurately train a than is presently possible.

With the proof of concept in hand with interest from several major industrial partners combined with a publication in the prestigious peer-reviewed scientific journal Physical Review Letters, the realization of Wolkow's life's work devoted to creating an economic way to scale up mass production of greener, faster, smaller technology is imminent.

"Initiating and monitoring the evolution of single electrons within atom-defined structures" appears in the October 15 issue of Physical Review Letters.

Explore further: Atomic-scale manufacturing now a reality

More information: Initiating and Monitoring the Evolution of Single Electrons Within Atom-Defined Structures, Physical Review Letters (2018). … ysRevLett.121.166801 ,

Related Stories

Atomic-scale manufacturing now a reality

May 23, 2018

Scientists at the University of Alberta have applied a machine learning technique using artificial intelligence to perfect and automate atomic-scale manufacturing, something which has never been done before. The vastly greener, ...

Writing the future of rewritable memory

July 23, 2018

Scientists at the University of Alberta in Edmonton, Canada have created the most dense, solid-state memory in history that could soon exceed the capabilities of current hard drives by 1,000 times.

Single Atom Quantum Dots Bring Real Devices Closer (Video)

January 27, 2009

( -- Single atom quantum dots created by researchers at Canada’s National Institute for Nanotechnology and the University of Alberta make possible a new level of control over individual electrons, a development ...

Recommended for you

CMS gets first result using largest-ever LHC data sample

February 15, 2019

Just under three months after the final proton–proton collisions from the Large Hadron Collider (LHC)'s second run (Run 2), the CMS collaboration has submitted its first paper based on the full LHC dataset collected in ...

Gravitational waves will settle cosmic conundrum

February 14, 2019

Measurements of gravitational waves from approximately 50 binary neutron stars over the next decade will definitively resolve an intense debate about how quickly our universe is expanding, according to findings from an international ...


Adjust slider to filter visible comments by rank

Display comments: newest first

not rated yet Oct 15, 2018
It looks like the future of AI will grow into an computational architecture no longer dependent on von Neumann and Turing machines. And we will see modeling of real phenomena, such as black hole collisions; and hypersonic airflow, at any scale, in any gas, that will accurately predict outcomes--99% of the time--by reason of neural network architectures for which we will grow to be accustomed to the fact that we won't have a clue how they actually arrived at their solutions.

This will lead to a scientific paradox: explanations of the underworkings of things for which the explanations themselves will be forever opaque, unless we want to take billions of years parsing the connections between octillions of electron-carrying molecules; and even then being dependent on the aid of gigantic classical computing engines with explicitly-tasked CPUs and memory, and hand-tuned commented code.
4 / 5 (1) Oct 15, 2018
DanR, right. Sooner than people think too, notice how they used machine learning to make the model, so AI building a AI architecture already.

"Forever opaque"? Maybe not, AI may be able to introspective and explain itself. It comes down to whether nature is indeed "too queer"
for us to understand as to whether we can follow the explanations.
not rated yet Oct 16, 2018
introspective and explain itself.

I'm sure it will be able to explain itself, to itself, and the fundamental elements and processes of neural networks are already well understood.
What will unlikely occur will be somewhat akin to mathematics' landmark 4-color map proof which required the massive use of computers, and even though the actual output depended on classical engines, the capacity of humans to parse it so that they actually understood it, could account for each step, and all the steps, was taxed beyond limit.

We are already at the point where neural network computing, again even dependent on von Neumann engines, can outperform--in certain areas--entire teams of human stock-market analysts, none of whom would be able to explicate how those machines arrived at their conclusions.

Last century saw the rise of quantum mechanics and its capacity for prediction. But we simply don't know how it actually WORKS; a precedent for a major scientific theory.

Please sign in to add a comment. Registration is free, and takes less than a minute. Read more

Click here to reset your password.
Sign in to get notified via email when new comments are made.