How the C12-CEA partnership can help drive quantum computing

French research institute CEA and startup C12 Quantum Electronics have announced a partnership to produce multi-qubit chips at wafer scale using carbon nanotubes. C12 and CEA also claim to be the world’s first to manufacture components to control qubits using standard manufacturing processes.

CEA says it manufactures quantum chips on 200mm silicon wafers by relying on standard CMOS processes, while C12 is pursuing carbon nanotubes as a novel approach to build qubits, which are the fundamental building blocks of quantum computers. Combining these two approaches would result in a scalable and ultra-coherent  platform for quantum computing. Coherency refers to the lifetime of a qubit: the amount of time it can retain its information before the quantum state is destroyed by noise. A full prototype is expected in 2024.

Addressing qubits using standard processes

Additionally, CEA and C12 claim to have demonstrated for the first time ever the capability to manufacture in volume core components to calibrate, control and read qubits using standard (presumably CMOS) processes. The nanotubes are assembled mechanically by C12 onto the semiconductor chip that has been fabricated by CEA. This allows C12 to design electronic circuits with almost arbitrary complexity, while protecting the qubit from contamination until the last manufacturing step, when the qubits and the silicon chip are combined. 

In this approach, the ultra-pure carbon nanotube serves as the qubit, while the silicon chip is used as a quantum communication bus. It also causes the qubits to be isolated. The carbon atoms are isotopically purified (which means that all atoms have the same amount of neutrons in the nucleus of the atom), which would minimize their decoherence by reducing noise.

“This partnership is a key milestone for our company to transfer an academic fab process to an industrial-grade semiconductor fab process, which was a major challenge,” said Pierre Desjardins, CEO and  co-founder of C12. “Thanks to CEA-Leti, we will benefit from better quality and higher volume as well as will prepare for industrialization of our devices.”

Finally, the two companies announced that they have started to manufacture chips for C12’s “quantum accelerators,”  which are meant for integration into classical supercomputers. This represents the startup’s first product milestone. 

C12 closed a $10 million seed round in June 2021.

More on quantum computing

Quantum computing represents a fundamentally new way to do computation, as it makes use of quantum mechanical properties such as entanglement and superposition. Whereas Moore’s Law is based on periodically doubling the number of transistors on a chip, in quantum computing the theoretical capabilities grow exponentially with each new qubit. 

Research has been progressing over the last decade across multiple companies and startups. Various techniques have been invented to make qubits. One of the earliest technologies that saw traction was the superconducting qubit. 

In contrast, C12’s qubits are created by trapping a single electron in a quantum nanotube. Gate electrodes are used to form a double quantum dot within the nanotube, and to entangle the electronic spin with the double quantum dot. The spin qubit is then addressed through a resonator via microwave pulses. Multi-qubit gates are performed via spin-spin coupling between the qubits.

Besides CEA, the main other company working on a CMOS-based quantum computer is Intel, which uses its 300mm fabs compared to CEA’s 200mm wafers. Intel is working on both silicon and silicon-germanium spin qubits, instead of carbon nanotubes, where a single electron is trapped inside a structure that resembles a classical transistor. 

Quantum applications

It is generally believed that it will require thousands or even millions of qubits for quantum computing to gain relevance for practical applications, which could go from chemistry to cryptography, materials science and optimization problems. With the industry currently stuck at around 100 qubits or even less, this means quantum computing still has some way to go before it could become a mainstream technology.
Nevertheless, some are already jumping on the bandwagon by offering the technology to the wider public for early adoption, for instance through simulators or cloud access. One such example is Azure Quantum.