And quantum supremacy was. Researchers working for Google claim to have reached this grail: demonstrate that a quantum processor is able to solve a series of computational calculations out of reach of supercomputers, even the most powerful in the world. It would take them more than 10,000 years to achieve this.
This unprecedented scientific breakthrough was described in an online article accessed by the Financial Times on Friday, Sept. 20, which was quickly removed from the Web. A disappearance that seems to indicate that the initial publication was a mistake or that it is an embargo that has not been respected.
More complex tasks done faster
But several copies of what seems to be the original article have appeared on the Internet for a week without Google researchers confirm the authenticity. Several specialists in quantum computing interviewed by France 24 believe that the searchable results appear serious and scientifically convincing.
"If this demonstration of quantum supremacy was to be confirmed, it is remarkable because it is an important step on the road to create a quantum computer", recognizes Iordanis Kerenidis, researcher at the Institute for Basic Computer Research (Irif ) The CNRS, contacted by France 24. The promises of the quantum computer are numerous: perform tasks more complex and much faster than a traditional machine. And the fields of application are multiple: medical research (development of new molecules for drugs), optimization of road traffic, artificial intelligence, etc. A powerful quantum computer should even be able to crack all the codes that are used on a daily basis, whether to encrypt electronic communications or protect sensitive data.
But so far, "traditional computers were able to perform all the tasks that have been submitted to a quantum machine," said Frederic Magniez, director of Irif, contacted by France 24. The results of research work of Google is changing the game: their quantum race beast has succeeded in just a few minutes to perform a task that would take several tens of centuries to the most powerful NASC supercomputers, say the authors of the article. Which amounts to saying that with "the current technology, this exercise is impossible to reproduce", summarizes Frédéric Magniez.
Having managed to cross the wall of quantum supremacy is only a first step. First, because the task submitted to this computational superpower "has been specifically designed to demonstrate the concept of quantum supremacy and has no concrete application," says Iordanis Kerenidis. According to the researchers interviewed, the next step is to replicate this feat with a "useful" task, such as, for example, modeling a molecule for new drugs.
Profits at the key ?
To achieve this, you may also need a more powerful machine than the one used by Google researchers. The latter used a computer equipped with a processor equipped with about fifty qubits, which are the equivalent in the quantum world of the traditional bits of our current computers. "What these researchers have used is a small quantum computer", recognizes Sébastien Tanzilli, project manager of quantum technologies at the CNRS Institute of Physics (INP), contacted by France 24.
This is not the type of calculator that could, for example, undermine the current cryptography systems, forcing the world to rethink all computer security. To do this, it would take a quantum computer with several hundred qubits. And "despite the actual technological progress, we do not know when such a computer will be developed, or even if it will ever exist," says Frédéric Magniez.
If Google and IBM seek to develop it, it's not just for the good eyes of science. "At first, there will not be a lot of quantum calculators, because it's expensive and difficult to do, so whoever has some of them can put them in the cloud and rent their services to customers who will need all this computing power, "says Iordanis Kerenidis. Customers with busy pockets such as pharmaceutical companies that would use it for their research, banks for issues of computer security or large cities wanting to better manage their traffic.
Although Google is probably motivated by the lure of profit, their demonstration of quantum supremacy is beneficial for the entire community, say the researchers interviewed. "If we did not manage to take this step, it would have made it more difficult to seek funding from industry, which should also motivate researchers to continue along this path," said Iordanis Kerenidis. For Frédéric Magniez, the whole industry gains in "credibility".