If the horizon of a quantum computer has not necessarily dissipated with the announcement of Google, and if some remain skeptical about the importance of this advance, the quantum promise still has its followers. The computer giants see it as a new battleground. Intel has a 49-bit chip (the conventional quantum bits counter, a computer measurement unit), IBM promises 53 for the end of the year, Google announced 72 in March ... IBM prefers to improve the quality of qubits rather than their number. Microsoft is exploring a qubit path of very long duration. All have already developed home programming languages.Article reserved for our subscribers Read also Google announces a major breakthrough in quantum computing
In their shadow, companies are making their mark on the market for quantum algorithms (QC Ware), business consulting (Quantum Benchmark), online systems to get their hands dirty (Rigetti, IonQ), investment funds ( Quantonation), simulators of these quantum computer-based machines (Atos) ... "We are probably the first to make money in this market," said Thierry Breton, the CEO of Atos, when of a special quantum day organized on June 20th by BPI France.
At their side, more unexpected companies are showing interest in this promising new technology. EDF, Total, Airbus, Bayer, Daimler and Merck already have research groups, monitoring activities or academic collaborations not to miss the quantum train, they hope for better numerical simulations, optimizations of certain processes, even discoveries of new molecules or materials ...
Moreover, the author of the term "quantum supremacy", John Preskill, pragmatic, develops the concept of "Noisy Intermediate-Scale Quantum" (NISQ, "noisy quantum system intermediate scale"). The acronym, which refers to the name of a classic computer architecture, RISC, refers to "small" machines of about 100 qubits, without error correction, but which could solve molecular modeling or materials science problems. . In other words, it's about seeing what can already be done with imperfect prototypes. Atos, which has been offering conventional supercomputers capable of simulating up to 41 qubits since 2017, plans to link such a NISQ machine with a supercomputer by 2023 in order to "accelerate" certain calculations.Article reserved for our subscribers Read also "Google, the ace of algorithms, the emperor of the Web, aims at quantum supremacy"
More precise measurements
The choice of technology has not yet been decided and this is where the key player in the field, namely academic research, still has a say, even if it has been overtaken by IBM, Google or Intel. The boil is still palpable and has even resumed since the announcement of several national investment plans in China, Europe or the United States (about 1 billion euros for the latter two over the next few years).
These means should be used to imagine ways, less hungry for qubits, to correct the fateful mistakes. But also to invent other qubits, the most advanced now, those of Google or IBM, are also the oldest. These "transmons" were invented in the 1980s by the French Michel Devoret and the American John Martinis, who took the lead of this research at Google, in 2014. In these circuits based on silicon, without electrical resistance and cooled to -273 ° C, the electrons behave as around an atom: they have only two possible states of energy. But other solutions already exist: teams trap atoms in lasers or electromagnetic fields or use grains of light as a quantum object.
And, if all these efforts of states, start-ups, industries do not work, quantum theory has other advantages than those of being a computational accelerator. Its properties promise more precise measurements of magnetic field, optics, electronics ... for ever more efficient sensors. They also promise safer communications. Calculation, sensor and safety are the pillars of the European investment plan in the field. In France, the scientific community is waiting for its national variation. A parliamentary report should come out at the end of November with proposals.Article reserved for our subscribers Read also Computer science: the big jump of the quantum computer