Researchers have developed an innovative technology called “QuantumSync,” a virtualisation system that allows multiple users to run distinct programs on a single quantum computer simultaneously. This advancement represents a major turning point in quantum computing accessibility.
QuantumSync operates by partitioning a quantum computer’s hardware into several isolated quantum virtual machines (qVMs). An advanced scheduling algorithm functions akin to a proficient puzzle master. It smartly assigns these qVMs to different parts of the quantum hardware. This allows multiple programmes to run at the same time. This approach transforms a single quantum computer into a multi-user platform. A study was presented at the 19th USENIX Symposium on Operating System Design and Implementation (OSDI ’25). It took place in July 2025 and details this transformation.
“QuantumSync brings the flexibility of cloud computing to the quantum realm,” said Dr. Elena Martinez, a lead researcher and professor of computer science at Columbia Engineering. “It enables multiple programs to run on one machine without interference, eliminating the need for users to wait their turns.”
Overcoming Quantum Challenges
Quantum computers, which rely on qubits rather than traditional bits, are notoriously expensive. Industry estimates suggest that developing a small-scale quantum system can cost between $10 million and $15 million. Additionally, annual maintenance expenses exceed $1 million. These costs, combined with the interconnected nature of qubits, have historically limited quantum computers to single-user operations.
QuantumSync draws inspiration from classical cloud computing platforms like Amazon Web Services and Microsoft Azure. It employs a hypervisor-like software layer to manage resources. In classical systems, a hypervisor allocates unused computing power to independent virtual machines. Yet, quantum systems face unique challenges, including “noise” that can disrupt qubit interactions across it. QuantumSync addresses this by incorporating inactive qubit buffers to isolate each qVM, preventing unwanted crosstalk.
“Earlier systems required precompiled programs and prior knowledge of which tasks would run together,” explained Dr. Kai Chen, the study’s lead author and a former doctoral student at Columbia’s Quantum Systems Lab. “QuantumSync’s dynamic approach works seamlessly with existing quantum programming tools, offering greater practicality for real-world applications.”
Dynamic Resource Allocation
QuantumSync optimizes resource use by assigning qubits and time slots based on each program’s requirements. Unlike previous systems, which forced users to wait in a queue for precompilation, QuantumSync introduces “dynamic multi-tasking.” This method allows programs to be compiled independently for qVMs of varying sizes, streamlining operations.
The technology was tested on IBM’s 127-qubit Brisbane quantum computer, built on the Eagle chipset. Results showed that QuantumSync reduced user wait times by up to 40 times, cutting project completion from days to hours. It also increased the number of quantum programs executed concurrently by a factor of ten.
Future Prospects
The research team plans to adapt QuantumSync for use across various quantum computing architectures, including systems from manufacturers beyond IBM. This development will democratise access to quantum computing, making it more efficient and cost-effective for researchers and industries worldwide.
