Pages

Wednesday, March 7, 2018

Speeding the advancement of quantum computing

By Rich Ptak

Google’s announcement of a 72-qubit chip has started a lot of noise about Google’s supposed
“quantum supremacy” and quantum industry leadership. We think it’s time to take a breath to understand a few terms and to discuss what’s real (and really available) in the world of quantum computing.

First, strictly speaking, the term Quantum supremacy is the potential ability of quantum computing devices to solve problems that classical computers practically cannot. Realistically, it is also used informally to describe definitively demonstrating/proving that quantum computers can outperform classical computers in problem-solving. It is about identifying the specific problem or class of problems for which quantum computer solutions are found find faster, more accurately, etc. than on a classic computer. That term most definitely does NOT simply refer to the chip with the highest qubit count.

Next, both quantum and classical computing solve problems using algorithms (step-by-step instructions taken to reach a solution). To date, efforts have failed to produce definitive proof of quantum superiority. Part of the problem lies in the limited ability to think quantum – that is, understanding and defining problems in quantum terms. One quantum computing developer who is helping in that arena is IBM, with IBM Q Experience (see below).

Next, let’s dispense with the idea that computing performance is simply related to the number of qubits. Theory says that a 5- or 16-qubit machine should be able to outpace any classical computer, i.e. more qubits equal more power. The NUMBER of qubits is less important than the QUALITY of the qubits. A 16-qubit high-quality array, i.e. stable, long-lived, controllable qubits is more powerful than a 100-, 500-, 2000-qubit device of not-so-stable, shorter-lived, inaccessible, low-quality qubits. IBM’s proposed quantum volume metric is based on such characteristics.

Finally, the real interest is in proving the UTILITY of quantum computing, i.e. the ability to actually do work. To that end, in 2016 the IBM Q Experience[1] introduced free access to quantum computing to the interested public (students, researchers, individuals, etc.). It offers a complete system including a developer’s kit (e.g. IBM’s QISKit[2]) not just a chip. Initially a 5-qubit processor, it added a 16-qubits processor in 2017. Participants include more than 1,500 universities, 300 high schools, and 300 private institutions worldwide. The results are impressive with in excess of 76,000 users running 2.9 million experiments and delivering more than 60 research papers.

Finally, for commercially oriented clients, there is the 20-qubit based IBM Q Network was announced in late 2017. Active participants include a dozen Fortune 500 companies, academic institutions, and national labs, in the list are JP Morgan Chase, Daimler AG, Samsung, Barclays, Honda, Keio University, Oxford University, University of Melbourne, and Oak Ridge National Lab. And a 50-qubit system is planned to be available by end-of-2018.

All this effort is to conclusively demonstrate quantum utility, not just quantum supremacy. Quantum computing will prove its value in implementation and application. That will happen as knowledge and the ability to think in quantum terms spreads into the community. IBM, with its open access to quantum systems, is helping that happen today. That is leadership.

You can find out more about IBM's activities in our  September 2017 write-up on Quantum computing and in the Featured Post (see column on the right) on the IBM Q Network. More is coming. 

No comments:

Post a Comment