It's not us, it's you: Boffins ditch supercomputers in lust for new materials

Tianhe-2, sit down. We've... we've built a simulator

AI Robot viewed from the back against an arty landscape. Pic via SHuttertock

Physicists from the National Institute of Standards and Technology have created a quantum simulator designed to model complex quantum mechanics phenomena that even the fastest supercomputer could not compute.

The researchers created the simulator by trapping up to 219 beryllium ions inside a Penning trap in a bid to study exotic materials, according to research published in Science.

A Penning trap is a device that holds the charged particles in suspension using a magnetic and electric field. The trapped ions interact with one another to produce quantum entanglement - a phenomenon required for quantum computers to work.

NIST attempted the same experiment in 2012, but did not manage to entangle as many ions and had more errors and instabilities compared to the new experiment. The new system has managed to scale up the amount of quantum entanglement generated to about 10 times as many ions as previous experiments. It can also be controlled better, say the researchers, making it more useful for studying quantum behaviour.

“Here we get clear, indisputable proof the ions are entangled,” NIST postdoctoral researcher Justin Bohnet said. “What entanglement represents in this case is a useful resource for something else, like quantum simulation or to enhance a measurement in an atomic clock.”

Ions have different spin states that act like quantum bits - or qubits - the unit of digital information used by quantum computers. By zapping the ions with lasers, the qubits interact and create a system that mimics the quantum behaviour of materials.

“Classical computers struggle to study such materials because entanglement means that the amount of information that must be stored exceeds even the memory of even the largest supercomputers today,” Justin Bohnet, lead author of the study, told The Register.

“Once you get to 30 to 40 particles, certain simulations become difficult,” Bohnet said. “That’s the number at which full classical simulations start to fail.

Co-author John Bollinger, who is also a researcher at NIST, said that this research “adds to a quantum tool box” that will be required to build a trapped-ion quantum computer in the future.

The main motivation for building a quantum simulator is to discover new materials that are related to "spin ices" and "spin glasses" - a class of materials that have interesting quantum behaviours due to their complex magnetic structures. It is also hoped that the new emergent behaviour could one day be used to develop quantum memories and quantum computers. ®

Sponsored: Becoming a Pragmatic Security Leader




Biting the hand that feeds IT © 1998–2019