Spinning Quantum

Quantum computing is a furnace for burning venture capital money. Yes, it will be transformative once we have useful quantum machines. But we are still very far off from any utility.

Quantum computing has a massive hype problem, as so many nascent technologies. Initially, there were grandiose claims of quantum supremacy, which turned out to be not only premature, but also bogus. Quantum supremacy was replaced with the more sensible notion of quantum advantage. Although no quantum computer ever achieved quantum advantage except on tiny toy problems, we are told to prepare for Q-day when encryption schemes will fail. Please bear in mind that the current record for factoring primes with Shor’s algorithm stands at 21 = 3·7. While larger prime numbers have indeed been factored on quantum annealers, we have no idea how these minimization algorithms actually scale, which means they may not offer any advantage at all. Furthermore, we appear to have already entered the era of quantum utility. Utility for whom, though? Clearly not customers, as evidenced by the $5m prize to find actual uses for current quantum computers.

That is hardly surprising. We are many years away from fault-tolerant quantum computers with enough qubits. Even a NISQ machine that is useful for a small class of realistic problems is still a while away.

Note that I am not talking about some sceptics who believe quantum computing is a hoax: if it is fundamentally impossible to build a quantum computer, they need to offer a mathematical proof, not merely hand-wavy arguments. After all, flying to and safely landing on the Moon was deemed impossible for millennia until 20 July 1969. The opinions of AI luminaries, without any credentials in quantum technologies, or researchers in somewhat related fields, who stop short of calling quantum computing a slush fund for postdocs, are also irrelevant. They all compete for the same pile of cash.

Winter is coming?

VC investments in quantum startups have almost halved from $2.2b in 2022 to $1.2b in 2023. The largest decline was in the United States. Investments in European quantum startups are growing slightly, though the outlook is hardly anywhere near “rosy”. A quantum computing winter is therefore not out of the realm of possibilities.

After HP (2019), PayPal (2022), UBS (2023), Alibaba (2023), and Baidu (2024) are shutting down their quantum computing labs. Alibaba and Baidu are noteworthy, because they both launched full-stack quantum platforms back in 2018 and 2021, respectively. Although the machine by Alibaba (Baidu) only had 11 (10) superconducting qubits, these are the first major cloud vendors backing down from quantum tech.

The few publicly traded, pure quantum technology companies have not been doing well on the stock market either:

• D-Wave is trading under $1.
• QCI dropped almost 99% to under a buck.
• Arqit is down 95% since 2021 and worth less than fifty cents per share.
• Rigetti cratered to 50¢ before going up again, though it is still down more than 85% from its IPO. The layoffs and executive turmoil had minimal impact on the market’s perception of the company.
• Only IonQ’s shares are roughly at their original level.

In 2022, revenue for the quantum industry was $942m. OMDIA expects a compound annual growth rate (CAGR) of 57.7% until 2032, which reads like the definition of a bubble. Tech companies that grow more than 50% annually and sustain that growth for a few quarters or even a couple years are not unheard of, but for an entire sector to do so for almost a decade is bonkers. By comparison, the hype around AI, fuelled by LLMs, is only expected to yield a CAGR between 15% and 19% over the next few years.

While almost a billion dollars sounds like a decent amount of revenue, there are between 80 and 350 quantum companies in the industry. With about 350 companies, the average annual revenue per company is only a bit under $3m. That is hardly enough to sustain large research labs without plenty of money from investors. Without a positive ROI or even a credible path to profitability, few investors, however, are going to care. The problem is not that the hype enables quantum computing companies to receive much-needed cash. The problem is that quantum computing cannot possibly live up to the current hype. Hype they are generating themselves.

Mostly by researchers for researchers

Few quantum hardware companies seem to understand that it is the entire ecosystem, not merely the technology that will ultimately determine success. Of course, without the unique quantum hardware, very little else matters.

Many quantum computers available are little more than laboratory equipment by researchers for researchers. Without quantum error correction integrated into the quantum stack, the knowledge required to solve even toy problems with acceptable error rates requires expertise that few companies outside the quantum industry are willing to invest in.

That is entirely expected: the hardware is not yet at the level where utility can or even should be expected. For instance, we do not yet know if there will be a single qubit modality that becomes dominant or if multiple modalities will coexist in different application niches. The problem is that companies are offering these machines to customers who expect some utility beyond the publication of a few research papers on problems that are mostly irrelevant to their core business.

The problem lies in the discrepancy between what quantum companies offer and what customers need. If I, as a customer, cannot solve a business problem more efficiently with a quantum computer, why should I bother at all? If I, as a customer with an R&D lab, cannot see a viable path towards solving business problems more efficiently on a quantum computer in the next few years, why should I bother right now?

Governments around the world have pledged $40b over the next ten years. Perhaps that is enough to sustain the industry while returning to a more reasonable level of buzz, as the industry seeks to make the technology useful for customers rather than merely available. If not, both VC and government funds will eventually dry up. Right now, it is all spin.