Google’s quantum computer failed to achieve quantum supremacy

Back in 2019, Google claimed it had achieved quantum supremacy with its quantum computer , but even then there were strong doubts about this claim. To deny what has been a research team, which was able to show how the Google statement on its quantum computer, Sycamore, that has reached the so-called quantum supremacy is, in fact, false.

Already in 2019, a few hours after the news leaked, IBM began to argue against this statement by questioning the result achieved by Google. And maybe they weren't all wrong.

quantum computer

What is quantum supremacy?

For the uninitiated, quantum supremacy is reached when you can perform a calculation on a quantum computer in acceptable times while a traditional supercomputer would take an incredibly exaggerated amount of time .

In this case it is a mathematical calculation that, according to what Google stated years ago, one of the best supercomputers in existence, Summit, would complete after 10,000 years. Google's quantum computer, based on the Sycamore quantum processor with 53 qubits, would have taken just 200 seconds.

The study, published on arXiv , reveals instead that by using a different algorithm than that used by Google in its demonstration, it is possible to obtain even higher performance than Sycamore using a traditional supercomputer .

quantum computer

The Sycamore quantum computer has not truly achieved quantum supremacy

To carry out the same calculation with a traditional supercompuer, in fact, it would not have taken 10,000 years as stated by Google, but only two and a half days:

"If our algorithm were implemented with high efficiency on a modern supercomputer with exaFLOPS performance, we estimate that ideally the simulation would cost a few dozen seconds, which is faster than that on Google's quantum hardware"

The new algorithm used is based on a different mathematical method than the one used by Google. This allows to arrive at a result very similar to that obtained with Sycamore, while running the algorithm on a traditional computer .

It is in fact possible to reduce the fidelity with which Sycamore's behavior is reproduced, still obtaining the same result and drastically lowering the time required for the calculation. It is obviously impossible to faithfully replicate the same quantum computer algorithm because it would require considering all possible alternatives , which requires significant computing power.

Given the obvious difficulty in using quantum computers, which at the moment are only scientific demonstrators and without practical applications, for some time we have stopped talking about quantum supremacy . The expression has in fact been replaced by quantum advantage : quantum computers are increasingly seen as machines that offer significantly higher performance than traditional computers in some computations of specific types, but not suitable to entirely replace classical devices .

According to scholars, quantum computers will one day be able to give birth to highly evolved artificial intelligences , and at the same time easily defeat modern cryptographic systems (see RSA) and, for this reason, not only companies, but also states themselves are doing of quantum computing a real priority.

The article Google's quantum computer failed to achieve quantum supremacy comes from Tech CuE | Close-up Engineering .