Predictions 2022: Quantum is not around the corner
A few more scientific and technical breakthroughs are needed before quantum computing starts becoming an industry. It might be five or twenty years ahead.
There was a time when technology hype was only instigated by marketing departments at tech vendor companies. Today it builds up already in scientific journals. Many technology vendors then take advantage of this to embellish their product descriptions and perspectives, creating further hype. Beware the trick. It is affecting technologies of all kinds, but especially life sciences and information technology.
The technology-industry complex has created the image of Quantum Computing as a readily available technology that just needs fine-tuning. But that is largely an illusion. If you talk to physicists, engineers and computer scientists who are not directly involved in quantum computing business initiatives, you will quickly learn that the situation is more complex.
Quantum information science is a marvelous discipline older than thirty years that some of the brightest people on earth are involved in. But the relevant technology, including quantum computing, is in its infancy.
Qubits are capricious objects with a tendency to disappear (‘collapse’) unless you keep them in an environment similar to that of interstellar space. Moreover, they tend to collapse as soon as you try to interrogate their status, which is the quintessential function of a quantum computer.
To go around such challenges, and many more, a half-dozen types of extremely different quantum computers are being tried. After five years of experimentation, none is clearly best, and this stalls the development of a quantum computing ‘industry‘.
And when an industry finally develops, we will be in for a disillusion: a lack of skills to make the technology work. Talk of AI writing quantum computing code instead of us is nonsense, given the required precision as well as the fact that we lack a sufficiently large data base for ML/DL algorithms to learn from.
More years will be needed to get the thing going. (For more, see Appendix).
Quantum computing is at the intersection of experimental physics and theoretical information science. We need a few more scientific and technical breakthroughs before it starts to become an industry.
This does not entirely rule out the possibility of a prodigious acceleration over the next few years, nor does it negate the promise of quantum technology as a powerful future accelerator for ordinary digital computers. But don’t count on it just yet.
So, exactly what is needed for an industry to develop and quantum computing to progress from possibility to probability? There can be more but I believe at least the following are the major road blocks:
- We have yet to see a quantum computer accomplishing a meaningful, practical task more efficiently than any classical computer.
- One, or perhaps a maximum of two QC architectures should emerge as clear winners. Otherwise, it would be hard for an ‘industry’ to develop. (The prospects are only partially mitigated by the fact that a quantum computer does not have to be general-purpose: all it takes is that a regular digital computer invokes its services). So far, quantum computers have been tried in analog and in pseudo-digital forms. Analog architectures include adiabatic, annealing, and direct simulation. Gate-based, i.e. ‘digital’ quantum computers, too exist in various forms. The most popular are ‘trapped ions’ and ‘superconducting qubits’. Both need careful environmental isolation including ultra-vacuum. The latter also require an ultra-frozen environment, while cooling of trapped ions can be obtained using special lasers.
- A logical qubit with a significant reduction in quantum error rate is a major milestone. Quantum Error Correction (QEC) enables noisy physical qubits to emulate stable logical qubits so that the computer behaves reliably. We don’t know how far QEC is although progress is reported in the scientific literature almost quarterly. And it would be just the first step to creating fully error-corrected gate-based quantum processors.
- All we have now are ‘noisy’ quantum computers (NISQs) and these are not demonstrably suitable to replace binary digital computers. They have been shown capable of some potentially reliable results via coupling with classical digital processors, developing hybrid quantum-classical algorithms. They work as follows. Since qubits are very short-lived in a NISQ, a quick ‘snapshot’ of a qubit state is taken in the quantum processor and is then processed using a classical computer. The classical computer then instructs the quantum processor on how to modify slightly the qubit state, and after this, another snapshot is taken. And so on.
- To date, any chosen implementation of a quantum computer requires the integration of technologies that are not present in classical, transistor-based processors production. Depending on the type of QC, such technologies can include ultra-vacuum, super cooling, mechanical super stability, special lasers, optical systems, microwave, coherent electronic controllers, and more. A path to a viable quantum computer must address these integration challenges, and this marks a divergence of the quantum computer from existing industrial infrastructures for computers.
Paolo Magrassi Attribution-ShareAlike 4.0 International (CC BY-SA 4.0)
I am the author of this article and it expresses my own opinions. I have no vested interest in any of the products, firms or institutions mentioned in this post. Nor does the Analyst Syndicate. I am indebted to Richard Marshall and Dan Miklovic for their suggestions.
From the same author: The asphyctic quantum industry (April 2021), Profiting from Quantum in five years? (February 2020), Quantum sweet dreams (December 2019), Quantum Computing? Sit back and relax (June 2019)