[Science] IBM says Google may not have reached quantum supremacy after all – AI

0
37
[Science] IBM says Google may not have reached quantum supremacy after all – AI


By Leah Crane IBM is in a race with Google to develop quantum computersCONNIE ZHOU/IBM The new dawn of quantum computing may not be as close as we thought. In late September, an unpublished paper from Google was leaked, claiming to have reached quantum supremacy – the point at which a quantum computer can perform calculations that would be impossible even for the best non-quantum supercomputers. But new work from researchers at IBM suggests those claims may not hold up. The Google research used a quantum processor called Sycamore that contains 54 superconducting quantum bits, or qubits, to check the randomness of a set of numbers. One of the qubits did not work, but the rest used their quantum entanglement to generate a set of ones and zeros and check that they were truly random. Sycamore did this in 3 minutes and 20 seconds. The Google team calculated that it would have taken IBM’s Summit, the world’s best supercomputer, 10,000 years. But according to new work from a team at IBM led by Jay Gambetta, Google missed something. Advertisement “Classical computers have such a large suite of things built into them that if you don’t utilise every single thing you leave yourself open for a tweaked classical algorithm to beat your quantum one,” says Ciarán Gilligan-Lee at University College London. Read more: Quantum computers are weirder and more powerful than we thought In this case, the IBM researchers say that Google did not fully take advantage of the supercomputer’s storage potential. Taking that into account, they calculated that it actually would be reasonable for a classical supercomputer to do this calculation. “An ideal simulation of the same task can be performed on a classical system in 2.5 days and with far greater fidelity” when the memory is taken advantage of, the IBM team wrote in a blog post. “This is in fact a conservative, worst-case estimate, and we expect that with additional refinements the classical cost of the simulation can be further reduced.” That means that Google did not actually demonstrate quantum supremacy, and the race is still on. Google did not immediately reply to a request for comment. This just goes to show that claims of quantum supremacy should be taken with a grain of salt, Lee says. “The type of quantum supremacy that Google and IBM are chasing here is really a moving target.” As quantum computers improve over time, so do classical computers – there’s always a new classical algorithm to test against the quantum ones, so the bar for quantum supremacy is continually raised. Reference: arxiv.org/abs/1910.09534 More on these topics: quantum computing