Building the next generation of supercomputers

New developments in super-powerful quantum computing and information processing at Yale are creating the foundations for the next technological renaissance.

Applied physics professors Michel Devoret and Robert Shoelkopf are working with a team of researchers to enhance the life span of quantum bits, revolutionary aluminum chips that will make computing much faster, they said. The advances have prompted the multinational technology and consulting firm IBM to increase its spending on quantum research, resparking innovation at the firm.

Three applied physics experts interviewed said the research is critical for the physical construction of a quantum computer containing the chips, as superconducting qubits (qubits that have been cooled to such low temperatures that their resistance to an electic current is basically nonexistent) show the most promise in harnessing quantum energy.

“This is critical research for actually building a quantum computer, as this directly influences the performance of single qubits as well as the complexity of algorithms that can be executed using the qubits,” University of California, Santa Barbara applied physics professr Andrew Cleland said. “There are potential work-arounds, but I don’t think they are realistic.”

While a classical bit is like a switch that can hold either of two positions, 0 or 1, Deveoret said, a quantum bit is more like a continuum which can hold multiple values ranging from 0 to 1. A quantum computer takes minutes to perform calculations that would take a regular computer thousands of years to do, such as factoring large numbers into their prime factors.

Qubits could potentially endanger and revolutionalize the field of data encryption, because their ability to process large volumes of data in minutes would make information transfer over the Internet extremely unsafe, Schoelkopf said. Unsurprisingly, he added, the National Security Agency is one of the largest motivators of qubit research.

A major advantage of qubits, Devoret said, is that they cannot be duplicated. If a qubit has to be transferred from one computer to another, the new qubit must be in a different form or the original must be destroyed. This natural phenomenon occurs because a qubit is destroyed in some way every time it is read.

“Today in the U.S., five percent of total energy is used in processing information,” Devoret said. “It is a question of saving energy for a society that could eventually need 10 times the present levels of energy. Qubits use the minimum possible amount of energy – they cannot work otherwise.”

In the lab, the team creates the qubits from aluminum metal before refrigerating the beam of microphotons and creating a very weak signal. This signal is then amplified and transferred to a monitor to be used in solving complicated algorithms.

People were pessimistic about the idea of quantum computing at first, Devoret said. The fragility of the quantum system coupled with the requirement of forcing billions of atoms and electrons to behave in the manner of an isolated atom made the reality of a quantum computer seem ludicrous. But today physicists have developed and worked with up to 12-15 qubits simultaneously, he said.

The Yale lab has created three qubits and is working on experiments with them simultaneously. They are attempting to study the external factors affecting the qubit, which has self-destructing properties, and hope to increase the time period for which the qubit can be sustained, postdoctoral associate Michael Hatridge said.

Experts in the field said the lab’s work shows promise.

“Their research has been instrumental in showing how powerful superconducting qubits can be and how important a candidate they are for quantum computing,” Kathryn Moler, associate professor of applied physics at Stanford, said.

Stanford physics professor Mark Kasevich added that the research is vital in demonstrating the viability of quantum computing schemes.

“With the current lifetimes of qubits, a quantum computer is impossible,” he said.

Schoelkof’s lab last published a paper in October 2010 called “High-Fidelity Readout in Circuit Quantum Electrodynamics Using the Jaynes-Cummings Nonlinearity.”