April 20, 2024

The cold, hard reality of quantum computing

The quantum computer revolution may be further away and more limited than many have been led to believe. That’s the message coming from a small but vocal group of prominent skeptics in and around the emerging quantum computing industry.

Quantum computers have been touted as a solution to a wide range of problems, including financial modeling, logistics optimization, and accelerating machine learning. Some of the most ambitious timelines proposed by quantum computing companies have suggested that these machines could be impacting real-world problems in just a few years. But there is growing pushback against what many see as unrealistic expectations for the technology.

LeCun from Meta: not so fast, qubit

Meta’s head of AI research, Yann LeCun, recently made headlines after pouring cold water on the prospect of quantum computers making a significant contribution in the near future. Speaking at a media event celebrating the 10th anniversary of Meta’s AI Fundamental Research team, he said the technology is “a fascinating scientific topic” but that he was less convinced of “the possibility of making quantum computers that are really useful.” .

While LeCun is not an expert in quantum computing, leading figures in the field are also sounding a note of caution. Oskar Painter, head of quantum hardware at Amazon Web Services, says there is currently “tremendous anticipation” in the industry and “it can be difficult to filter the optimistic from the completely unrealistic.”

A fundamental challenge for current quantum computers is that they are very error-prone. Some have suggested that these so-called “noisy intermediate-scale quantum” (NISQ) processors could still work usefully. But Painter says there is growing recognition that this is unlikely and that quantum error correction schemes will be key to achieving practical quantum computers.

“In the last 10 years we discovered that many of the things that people have proposed do not work. And then we found some very simple reasons for it.”
—Matthias Troyer, Microsoft

The main proposal involves distributing information among many physical qubits to create “logical qubits” that are more robust, but this could require up to 1,000 physical qubits for each logical one. Some have suggested that quantum error correction might even be fundamentally impossible, although that is not a widely held view. Either way, implementing these schemes at the scale and speeds required remains a distant goal, Painter says.

“Given the remaining technical challenges to realizing a fault-tolerant quantum computer capable of running billions of gates on thousands of qubits, it is difficult to establish a timeline, but I would estimate it is at least a decade away,” he says.

Microsoft: clarity please

The problem is not just one of time scales. In May, Matthias Troyer, a Microsoft technical fellow who leads the company’s quantum computing efforts, co-authored a paper in ACM Communications suggesting that the number of applications in which quantum computers could provide a significant advantage was more limited than some might lead you to believe.

“In the last 10 years we found that a lot of the things people have proposed don’t work,” he says. “And then we found some very simple reasons for it.”

The main promise of quantum computing is the ability to solve problems much faster than classical computers, but the exact speed varies. There are two applications where quantum algorithms appear to provide exponential speedup, Troyer says. One is to factor large numbers, which could make it possible to break the public-key encryption on which the Internet is based. The other is the simulation of quantum systems, which could have applications in chemistry and materials science.

Quantum algorithms have been proposed for a variety of other problems, including optimization, drug design, and fluid dynamics. But the touted speedups don’t always work: They sometimes amount to a quadratic gain, meaning that the time it takes the quantum algorithm to solve a problem is the square root of the time it takes its classical counterpart.

Troyer says these advances can quickly disappear because of the enormous computational overhead that quantum computers incur. Operating a qubit is much more complicated than changing a transistor and is therefore much slower. This means that for smaller problems, a classical computer will always be faster, and the point at which the quantum computer gains the advantage depends on how quickly the complexity of the classical algorithm increases.

Operating a qubit is much more complicated than changing a transistor and is therefore much slower.

Troyer and his colleagues compared a single Nvidia A100 GPU to a fictional future fault-tolerant quantum computer with 10,000 “logic qubits” and gates much faster than current devices. Troyer says they discovered that a quantum algorithm with a quadratic speedup would have to run for centuries, or even millennia, before it could outperform a classical one on problems large enough to be useful.

Another major barrier is data bandwidth. The slow operating speeds of Qubits fundamentally limit the speed at which classical data can be obtained in and out of a quantum computer. Even in optimistic future scenarios, this is likely to be thousands or millions of times slower than classical computers, Troyer says. That means data-intensive applications like machine learning or database searching will almost certainly be out of reach for the foreseeable future.

The bottom line, Troyer says, was that quantum computers will only really shine on small data problems with exponential speedups. “Everything else is a beautiful theory, but it won’t be practical,” she adds.

The paper didn’t have much impact on the quantum community, Troyer says, but many of Microsoft’s customers appreciated having some clarity about realistic applications for quantum computing. He says they have seen several companies reduce or even close their quantum computing teams, particularly in the financial sector.

Aaronson: welcome, skeptics

These limitations should not surprise anyone who has paid much attention to quantum computing research, says Scott Aaronson, a professor of computer science at the University of Texas at Austin. “There are claims about how quantum computing will revolutionize machine learning, optimization, finance and all these industries, where I think skepticism was always justified,” he says. “If people are just starting to accept that, then welcome.”

While he also believes that practical applications are still a long way off, recent advances in this field have given him reason to be optimistic. Earlier this month, researchers from quantum computing startup QuEra and Harvard showed that they could use a 280-qubit processor to generate 48 logical qubits, much more than previous experiments had achieved. “This was definitely the biggest experimental breakthrough in maybe several years,” Aaronson says.

“When you say quantum technology is going to solve all the world’s problems, and then it doesn’t, or it doesn’t do it now, that creates a bit of disappointment.”
—Yuval Boger, QuEra

Yuval Boger, chief marketing officer at QuEra, is keen to emphasize that the experiment was a laboratory demonstration, but believes the results have caused some to re-evaluate their time scales for fault-tolerant quantum computing. At the same time, he says they’ve also noticed a trend where companies are quietly shifting resources away from quantum computing.

This has been driven, in part, by the growing interest in AI since the advent of large language models, he says. But he agrees that some in the industry have exaggerated the technology’s near-term potential and says the hype has been a double-edged sword. “It helps get investment and gets talented people excited about getting into this field,” he says. “But on the other hand, when you say quantum technology is going to solve all the world’s problems, and then it doesn’t, or it doesn’t do it now, that creates a bit of disappointment.”

Even in areas where quantum computers look most promising, applications could be more limited than initially expected. In recent years, papers by researchers at scientific software company Schrödinger and a multi-institutional team have suggested that only a limited number of quantum chemistry problems are likely to benefit from quantum accelerations.

Merck KGaA: a lovely accelerator, sometimes

It’s also important to remember that many companies already have mature and productive quantum chemistry workflows operating on classical hardware, says Philipp Harbach, global group director of digital innovation at German pharmaceutical giant Merck KGaA, in Darmstadt, Germany (which does not should be confused with the American company Merck).

“In public, the quantum computer was presented as if it enabled something that currently cannot be achieved, which is inaccurate,” he says. “Primarily, it will accelerate existing processes rather than introducing a completely disruptive new application area. So here we are evaluating a difference.”

Harbach’s group has been researching the relevance of quantum computing to Merck’s work for about six years. While NISQ devices may have uses for certain highly specialized problems, they have concluded that quantum computing will not have a significant impact on the industry until fault tolerance is achieved. Even then, how transformative that impact could be really depends on the specific use case and products a company is working on, Harbach says.

Quantum computers stand out for providing precise solutions to problems that become intractable at larger scales for classical computers. This could be very useful for some applications, such as the design of new catalysts, says Harbach. But most of the chemical problems that interest Merck involve the very rapid detection of large quantities of candidate molecules.

“Most quantum chemistry problems do not scale exponentially and approximations are sufficient,” he says. “They are well-behaved problems, you just have to make them faster with a larger system size.”

However, there may still be cause for optimism, says Microsoft’s Troyer. Even if quantum computers can only address a limited range of problems in areas such as chemistry and materials science, the impact could still be game-changing. “We’re talking about the Stone Age, the Bronze Age, the Iron Age and the Silicon Age, so materials have a huge impact on humanity,” he says.

The point of expressing some skepticism, Troyer says, is not to diminish interest in the field, but to ensure that researchers focus on the most promising applications of quantum computing with the greatest potential for impact.

UPDATE: December 28, 2023: This story has been updated to clarify the fact that Matthias Troyer, quoted above, has seen a reduction in the financial sector but not in the life sciences sector, both of which Spectrum had originally reported wrongly.

From the articles on your site

Related articles on the Web

Leave a Reply

Your email address will not be published. Required fields are marked *