Senior Research Fellow and Senior Lecturer,
School of Computing and Information Systems, The University of Melbourne
Area of Science: Noisy intermediate-scale quantum technology, quantum algorithms
Current quantum computing technologies have limitations: quantum bits (qubits) don’t last anywhere near as long as classical transistors do, units of information are not transferred with absolutely 100 per cent accuracy, and noise can accumulate in circuits before an algorithm can complete, drowning out the answer. Dr Casey Myers is designing quantum algorithms to specifically work within these limitations – to demonstrate that our first generation of quantum computers can still do useful calculations faster than classical computers.
About Dr Casey Myers
Casey was studying physics at the University of Queensland when part of Australia’s first research centre for quantum computation and communication technologies was being established there. It put him in the right place at the right time to see the first steps being taken in the world of quantum computing, and led him to complete a PhD and then eight years of postdoctoral research in the field.
But the realities of life, location and occupation prompted a move out of academia and into ‘real world’ jobs. Theoretical physicists can use their skills in a wide range of areas, and Casey worked in seismic imaging for the oil and gas industry, and as a finance and risk analyst in the banking sector while continuing to follow research progress in quantum computing from afar. A chance to work in a quantum computing start-up company then gave him the opportunity to return to the field and catch up on the latest developments, letting him move back into full time research.
“When I left academia for industry in 2015, quantum computing was moving into a really exciting phase,” recalls Casey, “but machine learning hadn’t really made a mark on physics research yet, and access to small quantum computing systems meant caging favours from the experimental physics team working in the basement. Now I’ve come back to a research world where access to physical quantum computers exists, and cloud access to a small diamond-architecture quantum computer and HPC-scalable quantum simulator will be available at Pawsey next year. I can now log in and put my quantum computing algorithms into the queue to test how they will perform on actual quantum-based hardware. It is a fantastic capability that is really progressing our knowledge of what quantum computers may eventually be able to do.”
What drew him to science?
“In high school, I just started reading science books. I came across one of Stephen Hawking’s books, and it led me to ask questions about the Universe in general, and find out what theoretical physicists like him actually do. By the time I finished school, I was absolutely sure I wanted to study physics, because that was the way to get the questions answered that I was really asking.”
Research with supercomputers
‘Quantum supremacy’ – demonstrating that a quantum computer can solve a problem faster, more efficiently, or more accurately than a classical computer – has already been achieved. But Casey notes: “These demonstrated problems are fairly contrived, they’re designed for the way quantum computers work. To make quantum computing a reality we need to demonstrate ‘quantum advantage’ – solving a useful, practical problem faster or more accurately than any classical computer.”
Casey is using Pawsey supercomputing to run his experimental quantum algorithms on the Quantum Brilliance Quantum Emulator, which uses classical computing to simulate how a quantum computer would operate. The aim is to design the algorithm so it can deal with the current physical limitations of quantum computing hardware, primarily the introduction of noise and errors during the time it takes to perform a calculation.
Casey is working on two complementary approaches to designing quantum algorithms: reducing the size of the algorithm so it can execute before being overwhelmed by noise, and incorporating quantum error mitigation techniques to limit the introduced errors and recover the quantum advantage that would otherwise be lost to noise.
Real world solutions
The problem Casey is working to solve is drawn from the world of finance – working out the right price for an option on the stock market. “Using the quantum emulator I can see how these programs would run, to get them working and see if they can perform faster than a classical computer.”
Once Quantum Brilliance’s diamond-based quantum hardware is installed at Pawsey next year, he’ll be able to queue up his algorithms over the cloud and test them on the real thing. “That’s the ultimate goal, to see these algorithms run on a quantum computer, and beat a classical computer with a practical problem. Even if we only show a low-level polynomial improvement in speed, it’s significant. If we can cut the run time or increase the accuracy of even simple problems, the intractable problems that have too many calculations for even our supercomputers to handle start becoming approachable.”