Researchers at the French Alternative Energies and Atomic Energy Commission (CEA) in Grenoble are confident of reaching a key milestone at the end of this year in their quest to build a quantum computer.
Maud Vinet and Silvano De Franceschi from the CEA along with Tristan Meunier of CNRS are leading a multidisciplinary team of around 50 scientists and engineers to build a silicon based quantum machine, the first critical step of which would be to operate a network of two qubits in the coming months.
How quantum computing works
Qubits are the units of information in quantum computing. They are the quantum equivalent of bits. Unlike classical computing where bits can exist as either 0 or 1, in quantum systems they possess both values at the same time. This property is called superposition.
The other key quantum property is called entanglement. It refers to the almost instantaneous effect two qubits have on each other even at a distance after having been initially coupled. Entanglement and superposition give quantum computers their phenomenal calculating power.
But keeping qubits entangled is a big challenge. “It is subject to interference from the environment. Any disturbances, whether thermal, electrical or mechanical, can cause errors,” De Franceschi says.
One way to limit the errors caused by these factors is to operate the qubits in a deep freeze mode.
“When qubits are cooled down to sufficiently low temperature, typically below a few degrees Kelvin, they are no longer susceptible to undesirable thermal excitations and their coherence can be preserved,” Vinet says.
Though the system to cool qubits uses the similar principle as that of a household refrigerator, it is much bigger and way more complex.
The CEA has several cryostats that use helium to achieve a temperature between 15 millikelvin to 1 Kelvin.
That corresponds to 272 degrees below water’s freezing point. Besides the above-mentioned cryostats, the CEA also boasts of a cryogenic prober that can carry out automatic measurements of 300 mm silicon wafers below 2 Kelvin or minus 271 degree Celsius. There are only two such machines in the world.
The French approach
There are four major approaches to fabricate the qubits: photons, trapped ions, superconductors and semiconductors like silicon.
Vinet and De Franceschi have adopted the last approach which involves the use of the magnetic moment of an electron in silicon to create the two different states of the qubit. They have chosen silicon even though it seems to be lagging behind the others in terms of the number of interacting qubits in a network.
“The other three approaches seem to have made more progress. But we are sticking with silicon. That’s because building workable quantum computers is not a short term race. It doesn’t matter where you stand today. What matters more is the growth potential for the future,” De Franceschi told RFI.
According to Vinet, in order to build a practical quantum computer, scalability will be the key. “In this regard, there’s no better candidate than silicon, which is central to the semiconductor industry. With silicon we can fabricate millions or even billions of qubits that can be assembled in a relatively compact system. It’s also convenient for control electronics.”
Moreover, according to De Franceschi, when it comes to performance, the silicon qubits are on par with the other platforms in terms of fidelity and speed of operations. De Franceschi contends that some of the other approaches may be appropriate but they may not be equally suitable when it comes to effective, massive and easy manufacturing.
“You need to consider how good you can scale up and handle the controlling of qubits once the processor size grows. There are other problems such as possible interference when you are manipulating qubits. The successful approach will be the one that copes the best with all these issues,” he says.
Researchers at CEA have a unique advantage as both the physics and the engineering requirements necessary to build a quantum computer are available under the same roof.
While De Franceschi and his team are engaged in perfecting the fabrication and interactions between qubits, Vinet and his group are working in parallel to make qubits truly scalable and to build the other components of a quantum computer.
“What we are trying to do here is build a full stack quantum computer. We are developing the quantum chip, control electronics, implementation of the quantum algorithm as well as an interface that translates the algorithm into electrical signals,” Vinet says.
Quantum computers have elicited huge interest from not just research labs but also IT giants, start ups and governments. In January 2021, French President Emmanuel Macron announced a 1.8 billion euro Quantum Plan initiative for supporting research and development of quantum technologies.
The enormous appeal of quantum computing lies in its promise to easily outperform even the world’s most powerful supercomputers on certain types of calculations.
“They are expected to solve complex problems such as protein simulations, calculating air flow on aircraft, finding new materials such as possibly room temperature superconductors,” Vinet says, adding researchers still don’t know how powerful these machines will turn out to be.