The Engineer – Commentary: Quantum computing has a hype problem
Peter Debenham, senior consultant at Plextek, looks behind the headlines and explores realistic future applications of quantum computing
As a buzz phrase, quantum computing lives up to the likes of AI, IoT, and the Metaverse in hype. But ask people what quantum computing is and what it means to have quantum computers for the future, and most will struggle. If you dig a little deeper, the only thing that sticks in people’s minds are the headlines about how quantum computers are threatening to destroy internet security. If quantum computers are fast enough to crack encryption algorithms, it’s game over. But is it a reality?
First, quantum computers are sufficiently in their infancy that, in most practical cases, they do not currently exist. It wasn’t until 1980 that physicist Paul Benioff proposed a quantum mechanical model of the Turing machine, and scientists and engineers are just beginning to develop the physics and engineering needed to build commercial quantum computers. Larger processors contain hundreds of qubits, the basic unit of a quantum computer, where tens of thousands or millions are needed. The problems of building quantum computers are not insurmountable but will take more time.
An article from 2021 suggests that with just over 13,000 qubits, a quantum computer could factor a 2048-bit RSA integer; the kind of operation needed to break much of the existing internet encryption. They wanted 177 days to do it, but that’s much faster than the “not before the universe heats up” answer for typical computers. Another article from 2021 suggests that 20 million qubits would only need 8 hours to do the same thing.
Most of those working in cybersecurity have been aware of the outstanding issues for many years and have developed cryptography resistant to quantum computers. In 2015, the US National Security Agency (NSA), followed by the UK National Cyber Security Council (NCSC), announced their intention to move to quantum-resistant cryptography and have standardized algorithms ready for 2022-2024. NIST, the US National Institute of Standards and Technology, announced four candidate algorithms on July 5, 2022.
What are they for?
Given that quantum computers suddenly won’t allow everyone to read the world’s encrypted internet traffic, what good are they? Quick search and machine learning problems, yes, but what else?
Two physicists, Manin in 1980 and Feynman in 1981, answered this question by pointing to simulating things you can’t simulate with classical computers, such as quantum mechanical systems. We can accurately model the quantum mechanics of simple systems like a single hydrogen atom using pen and paper and a few particles using existing computers. But larger quantum mechanical systems cannot be modeled accurately at all without incredible simplification.
Feynman imagined trying to use quantum mechanics in our current computers to accurately model a system with a larger number of elements. For a number of particles, N, at a similar number of positions in space, you need memory to store and processing to compute NNOT configurations, which quickly become too large. Even for 100 particles you have about 10200 configurations to store and calculate at each step. Compare that to estimates of 1080 atoms in the observable universe and that is impossible for classical computers. But this could be done with a quantum computer of the same order of magnitude of qubits, namely 100, because the qubits react in the same way as the modeled system.
Existing simplified quantum mechanical models underpin modern chemistry, materials design, and pharmaceuticals. Perfectly accurate models would allow much more. New drugs, more efficient chemical processes and new materials.
For example, the first production of fertilizer using the Haber-Bosch process consumes about 1% of the world’s total energy production and generates 1.4% of the world’s CO.2 generation. Fertilizer is needed to feed a world of 8 billion people, but better modeling gives the high possibility of a more efficient process by designing a better catalyst.
The second example is material design. To decarbonize, the world is moving from internal combustion to electric motors. A huge problem with motors is heat from electrical resistance. Heat is not only a loss of efficiency but worse, there is the problem of how to dissipate it. If it’s too hot, the engine fails or something catches fire. Less heat allows for smaller, more efficient motors that are quieter too. Replacing motor wiring with high temperature superconductors eliminates both problems, but applications are limited because for superconductors high temperature means above liquid nitrogen (77K or -196ohVS). Better material modeling would facilitate the search for useful room-temperature superconductors, moving superconducting motors from large industrial environments to normal everyday life.
So what will quantum computers do for us? It won’t break internet security, but it might give us better chemical processes and room temperature superconductors. Quantum computing could change the world, but right now its future remains uncertain.
Peter Debenham is a Senior Consultant at Plextek
 Gouzien, E. and Sangouard, N., 2021. Factorization of 2048-bit rsa integers in 177 days with 13,436 qubits and multimode memory. Physical examination letters, 127(14), p.140503.
 Gidney, C. and Ekerå, M., 2021. How to factor 2048-bit RSA integers in 8 hours using 20 million noisy qubits. Quantum, 5p.433.
 Alagic, G., Alperin-Sheriff, J., Apon, D., Cooper, D., Dang, Q., Dang, T., Kelsey, J., Liu, YK, Lichtinger, J., Miller, C. , Moody, D., Peralta, R., Perlner, R., Robinson, A., & Smith-Tone, D., 2022. Progress Report on the Third Cycle of NIST’s Post-Quantum Cryptography Standards Process . US Department of Commerce, NIST.
 Manin, YI, 1980. Vychislimoe i nevychislimoe (computable and non-computable), Moscow: Sov.
 Feynman, RP, 1982. Physics simulation with computers. International Journal of Theoretical Physics, 21(6/7) (publication of a conference talk from May 7, 1981)