Google is pushing quantum computing closer to reality

By Thomas Black

Alphabet Inc’s Google has reanimated excitement over quantum computing with an announcement about how its new chip, Willow, trounced a classical computer to solve a mathematical equation much faster. More important, Google said it had crossed a crucial threshold on the way to an error-free quantum computer, the holy grail of the technology.

The tech world has been down this hype alley before. Google instigated that round with the 2019 introduction of its Sycamore processor, which also solved a mathematical equation that led the company to claim quantum supremacy over classical computers. News articles exhausted every angle in describing how these magical machines were going to impact each industry in the future. That enthusiasm fizzled out because while the machines were proven, they were not ready for widespread use. Besides, Nvidia Corp. keeps making superpowerful chips that are pretty good at simulating quantum computing without the technical drawbacks.

This is where it is helpful to understand the difference between classical and quantum computing. Classical computers use digital bits that process information in 1s and 0s, made possible by electric current that turns on or off tiny transistors. Quantum computing uses, well, quantum bits, or qubits, which are able to process data between 1 and 0 using natural or human-made particles. Using a paint analogy, the bits on a personal computer chip are the equivalent of black or white. Qubits, though, can come in all shades on the color wheel. That potential array of values is what makes quantum computing so powerful and also underpins the reason the machines will be able to do much more complex calculations than their black-and-white classical cousins. The problem is that qubits are extremely delicate and can cause quantum computers to lose information, making them prone to error. Google’s latest achievement is that it can correct errors at greater scale.

It’s a good thing that Google’s announcement has perked up ears again on quantum computing because the industry has been quietly making great strides toward creating machines that researchers from companies and government find valuable, and investors should exercise patience. Some computer makers, like IonQ Inc, say their machines are already providing practical results. The startup, whose shares have jumped more than threefold this year, is planning to manufacture five quantum computers at its factory and just set one up in Switzerland.

Many models of these early, error-prone computers, including machines by IonQ and Rigetti Computing Inc, can be accessed both through Amazon.com Inc’s Braket and Microsoft Corp’s Azure. Other quantum computers available on these portals are produced by Pascal, Quantum Circuits Inc, and Quantinuum, which is controlled by Honeywell International Inc. International Business Machines Corp. provides software tools and access to several of its models through Qiskit. Google provides software tools and simulation but not general access to its computers.

In other words, functioning quantum computers, albeit prone to error and limited, are available through cloud-based networks. Industry stalwarts, such as California Institute of Technology’s John Preskill, are getting excited after years of repeating the same mantra that quantum computing was a decade away.

“The quantum hardware has reached a stage now where it can advance science,” Preskill said in a video accompanying the Willow announcement. “We can study very complex quantum systems in a regime that we’ve never had access to before.”

The systems are getting better and better, Preskill said. That’s why Honeywell, which owns 54 per cent of Quantinuum, should resist a push by investors to monetize its stake in the quantum computing startup. The industry is at the cusp of providing a useful tool to supercharge research across industries, and the value will only rise. That gets lost in the argument from investors to simplify Honeywell’s conglomerate model.

Again, quantum computers still aren’t ready for prime time because the machines are prone to error. But the race is entering the final stretch on who will build the first practical quantum computer, making this a great time to start paying attention for investing opportunities, not to mention the pure entertainment value. This is a real-life reality show that pits teams of scientists against one another to become the founders of a new era of computers. In the end, it may be hard to declare a winner because the gains are likely to be incrementally better until the impact on scientific research becomes apparent.

What makes this race so entertaining is that it pits two broad technological pathways, or camps, against each other to reach the ultimate goal: a machine with enough error-corrected qubits to do advanced calculations. The divide between these camps will answer this critical question: Can humans fabricate an object that creates a quantum state that can be tweaked and tuned enough to match the natural quantum state provided by particles such as atoms or photons?

On the manufactured-qubit side are Google, IBM, Rigetti, IQM and other companies, which are building computers with superconducting qubits. In its Willow announcement, Google discussed its fabrication facility in Santa Barbara, California, that’s built especially for making superconducting qubits and how it improved drastically the time its qubit remains in a quantum state.

On the other side, companies that create qubits from natural particles, such as atoms or photons, believe that the qubits manufactured on techniques based on semiconductor production advances will hit a wall in their ability to make exact qubits and connect them. This camp —which includes makers of trapped-ion computers such as IonQ and Quantinuum — instead capture, control and manipulate atoms (or photons for some startups) with lasers. The superconducting camp points out that having to move around atoms and other particles with lasers introduces errors and slows down computing speeds. The ability to scale is a huge hurdle for computers built with natural particles, whereas the qubit manufacturers are supported by a semiconductor industry that has already scaled.

Qubits, though, aren’t all created equally. Remember, it’s real particles versus human-made ones. Perhaps both pathways will have a place in the market. The industry standard will gravitate toward how many error-corrected, or logical, qubits a machine makes available to coders. In its Willow announcement, Google announced it had surpassed the “below threshold” at which it can add qubits and reduce errors. This is important because quantum computers need to add spare qubits to correct and maintain the logical, or error-corrected, qubits on which computations are made. Classical computers are also corrected for errors, but the fault rate from misfiring transistors is tiny to begin with.

The makers of computers based on atoms — Quantinuum, IonQ, Atom Computing and others — think they are ahead in the race because their qubits have lower error rates. Quantinuum and Microsoft Corp. published a paper in April that detailed how four logical qubits were created by using 30 physical qubits. Keep in mind that researchers have said that even a computer with 100 error-corrected qubits would allow for calculations unmatched by classical computers.

The long race toward quantum computing is entering the backstretch, and the progress is accelerating. There will be business opportunities. IonQ’s shares just hit an all-time high. Rigetti’s shares have more than doubled just since Dec 6. Quantinuum will likely sell shares to the public in the near future. Google has amped up the excitement. Investors like those in Honeywell should realize that the payoff will be worth the wait.