The new system turns quantum bottlenecks into breakthroughs

Quantum computers operate under major limitations. You can only run one program at a time. These million dollar machines require exclusive use even at the smallest tasks, leaving many of the expensive and fast hardware idles, forcing researchers to withstand long queues.

Researchers at Columbia Engineering have developed HyperQ, which allows multiple users to share a single quantum computer simultaneously with an isolated quantum virtual machine (QVM). This critical development brings quantum computing closer to real-world ease of use. This makes it more practical, efficient and more accessible.

“Hyperq brings cloud-style virtualization to quantum computing,” says Jason Nee, professor of computer science at Columbia Engineering and co-director of the Institute of Software Systems. “You can allow a single machine to run multiple programs at once. It’s not interference, you don’t wait in line.”

Overcoming basic bottlenecks

This approach shows a dramatic change from the traditional one-time model. By dynamically allocating quantum resources and intelligently scheduling jobs, HyperQ analyzes the needs of each program and directs them to the best part of the Quantum chip, allowing multiple tasks to run immediately without slowing down each other.

Just like how Cloud Servers revolutionized classical computing by maximizing efficiency and scalability, HyperQ aims to bring the same transformation possibilities to quantum computing, allowing for wider access, faster turnaround times and more productive use of limited quantum resources.

Directed by Nieh and Ronghui Gu, associate professors of the Tan Family of Computer Science in Columbia, the new work is scheduled to be released on July 8th in Boston, July 7-9, 2025 at Usenix Symposium on Operating System Design and Implementation (OSDI ’25).

A new layer of quantum control

HyperQ is a software layer, hypervisor, inspired by virtualization technology that powers modern cloud computing. Split the hardware of a physical quantum computer into multiple small, isolated quantum virtual machines. The scheduler then acts like a master tetris player, packing multiple of these QVMs and running them simultaneously on different parts of the machine.

Researchers ran a HyperQ prototype on IBM’s largest quantum computer via IBM Quantum Cloud. It is the first time to introduce and implement the concept of virtual machines for multiplexing actual quantum computer hardware, beware of researchers.

“Previous efforts required specialized compilers and needed to know exactly which programs would run together in advance,” said the lead author of The Paper and formerly at the Institute of Ph.D. Software Systems, Columbia. “Our approach works dynamically with existing quantum programming tools. This is much more flexible and practical for practical use.”

The system reduced the average user latency up to 40 times, converting turnaround times from day to minute hours. Additionally, the number of quantum programs running in the same time frame increased by up to 10 times, ensuring a much higher utilization rate of expensive quantum hardware. Surprisingly, HyperQ’s intelligent scheduling can even increase computational accuracy by steering sensitive workloads from the noisiest areas of quantum chips, Gu pointed out.

Wide range of impacts across the industry

Hyperq’s success potential is broad. For quantum cloud providers like IBM, Google, Amazon, and more, this technology provides a powerful way to serve existing hardware infrastructures to more users, increasing both capacity and cost-effectiveness.

For academic and industry researchers, HyperQ means access to quantum computing resources is much faster. This acceleration could dramatically speed up work in key areas such as drug discovery, advanced materials development, and the creation of more efficient energy solutions. Ultimately, these advancements can bring great benefits to society, from improved medical outcomes to more sustainable technologies.

Looking ahead

Nieh and his team plan to extend the capabilities of HyperQ to support new quantum computing architectures and to provide enough flexibility to work with any type of quantum computer. That way, as quantum technology evolves, HyperQ can continue to help many users share quantum hardware efficiently, regardless of the underlying technology.

A new era of quantum computing

HyperQ represents a crucial step in transforming quantum computing from powerful yet constrained scientific tools to practical technologies that are poised and sought-after to highly foster real-world impact.

“Instead of forcing one person to monopolize the entire machine, many users can now share quantum resources at once at once,” says Tao, now an assistant professor at the University of Maryland College Park. “This will change the game of how quickly we can tackle some of the world’s most challenging issues.”