How to scale-up cryogeny for large-scale quantum computing ?
One of the requisites to harness the full potential of quantum computers is to be able to increase the number of qubits they operate. Several companies, such as Google or IBM, have set a clear goal to reach computers comprising about one million qubits within the next decade, where the maximum number of qubits today is several dozens at best.
Many qubits technologies have to operate at ultra-low temperature, sometimes as low as 10 milliKelvin (0,01 degree above absolute zero). This extreme cold is necessary to preserve the coherence of qubits that would otherwise be too prone to error to be of any practical use.
The matter is that as the number of qubits increases, more and more heat load will be brought to the cryogenic system, thus posing a clear threat to its ability to stay at low temperature. To get to a thousand qubits, which is the next milestone for the technology, an increase by at least a factor 10 is needed at ultra-low temperature, and by a factor 100 at 4 Kelvin. It becomes clear that to reach a million qubits, a simple increase in the number of cryogenic systems will become impractical.
Hopefully, by rethinking the cryogenic architecture used for quantum computing, a large leap could be achieved. For instance, the smallest helium liquefiers routinely produce the same cooling power at 4 Kelvin as a dozen of cryocoolers, and the largest plants exhibit powers that would be far enough for large-scale quantum computers. This would imply to rethink the way cold power is used all the way from room temperature to absolute zero, so that each stage works at its most efficient. As an example, liquid nitrogen could provide a first stage of thermal shielding at 77 Kelvin, then liquid helium to go down to 4 Kelvin, and dilution refrigerators would achieve the last step close to absolute zero. Such a thermal staging is extremely classical and enables each technology to work at its best. In past decades, Air Liquide has demonstrated its ability to build such large and complex systems to answer the needs of experiments such as ITER or CERN. Together with Cryoconcept, acquired in 2020, it now has all the bricks to work on integrating the cryogenic bricks to make a quantum computer.
As a consequence, it will be necessary to rethink all the electronics within a quantum computer so that it works hand in hand with the cryogenic system. This work can’t be done alone, and Air Liquide is looking for partners to define the next cryogenic platform for quantum computing.