For years, quantum computing has been pitched as one of the next great leaps forward in technology. A 2026 McKinsey analysis suggests this new approach to complex problem solving could unlock nearly US$2.7 trillion in value by 2035, opening the door to breakthroughs in areas ranging from finance and transportation to new drug discovery, more sustainable materials, more efficient supply chains, and stronger cybersecurity.
Interest in the technology is accelerating, with investment in quantum technology start-ups reaching US$12.6 billion in 2025—more than six times the year before. Ninety percent of that money went to quantum computing.
But for all the excitement, big questions remain about what it will take to move quantum computing from the research lab into practical use. Those challenges took centre stage at last year’s IEEE Quantum Week conference in Albuquerque, New Mexico. Delivering the keynote was Sam Stanwyck, director of quantum product at Nvidia, who described the field as reaching an inflection point.
“The era of logical qubits has arrived,” he told attendees, pointing to advances that are helping to make quantum systems more reliable. But he cautioned that the technology will only deliver meaningful impact if progress across hardware, software, and error correction comes together to create systems capable of solving problems at scale.
“There’s a lot of work to do,” he said.
Bigger doesn’t mean better
For Stanwyck, one of the biggest misconceptions about quantum computing is that there will be one defining moment that suddenly makes the technology useful. In reality, progress depends on solving several difficult engineering challenges at once—building hardware that can perform more calculations with fewer mistakes, developing software and algorithms capable of tackling the kinds of large, complex calculations quantum systems are meant to solve, and making systems reliable enough to work at scale.
“Last but certainly not least, we need to solve scalable quantum error correction,” he said.
That final hurdle may be the toughest. Quantum systems are notoriously fragile, meaning even tiny disturbances—heat, stray signals, or even just the passage of time—can introduce errors into calculations.
Today’s leading quantum computers still operate with error rates of roughly one in every thousand operations—a frequency Stanwyck called “incredible,” and enough to enable some significant solutions. “But for the applications that we’re most excited about, that we know have large quantum advantages, we need many, many more orders of magnitude decrease in error rates,” he said.
He also urged caution about how progress gets measured. The number of quantum bits (or qubits, the basic unit of quantum information) is often taken as a sign that systems are getting more powerful, but bigger numbers don’t always mean better machines, explained Stanwyck. He pointed to a recent example in which researchers simulated a system of roughly 65,000 qubits—impressive, until it became clear that the qubits barely interacted, with calculations running only briefly before errors began.
That reality is already pushing some companies toward more tightly connected hardware, rather than simply boosting qubit counts.
Accelerated supercomputers
Still, Stanwyck said promising advancements are happening, even if some of the biggest technical hurdles remain unresolved. One of the biggest shifts is a growing recognition that the technology will not replace today’s computers—at least not anytime soon.
“Quantum computers will not stand alone,” he said. Instead, he described a future where quantum systems work alongside classical computing, with CPUs and GPUs helping to handle everything from simulation and preprocessing to the heavy lifting required for error correction.
“We are constantly seeking to expand the set of problems that computers can solve and do it in a way that helps the whole rest of the world be more successful.”
Stanwyck called that approach a “quantum-accelerated supercomputer”—a tightly integrated system designed to tackle problems neither classical nor quantum computing could solve on its own.
For Nvidia, best known for the chips powering today’s AI boom, that means focusing less on building quantum computers themselves and more on the infrastructure around them. Stanwyck said the tech giant sees itself as an “accelerated computing platform company,” creating tools that researchers and tech innovators can use to experiment, test ideas, and move the field forward.
“We are constantly seeking to expand the set of problems that computers can solve and do it in a way that helps the whole rest of the world be more successful,” he said. The goal, he added, is to “build the tools to make it easy to use for everyone.”
Stanwyck compared the approach to Nvidia’s early investments in GPU computing and deep learning, when the company focused on building software and tools before the technology became mainstream.
“We’ve been working on these technologies in the time before the big breakthroughs, trying to create the conditions for those breakthroughs,” he said.
That sense of optimism is likely to shape discussions when IEEE Quantum Week returns Sept. 13 to 18 at Toronto’s Metro Toronto Convention Centre, bringing together researchers, engineers, startups, and major technology companies to assess where the field stands and what needs to happen next.
For all the challenges ahead, Stanwyck sees signs that quantum computing is finding its footing.
“We’re making great progress with all of you, and excited for what comes next.”
PRESENTED BY
Learn more about IEEE Quantum Week 2026.
Feature image courtesy IEEE Quantum Week.