Quantum middleware: the glue between bits and qubits

Quantum computers will most probably revolutionize computing. However, the future of quantum computing largely depends on its integration with classical computing, as a complete shift to quantum will probably never happen—and certainly not overnight. The foreseeable future will be hybrid, where quantum and classical hardware and software systems coexist within one framework. Together with collaborators, such as BMW and Princeton University, computer scientist Nishant Saurabh is developing quantum middleware to bridge these two worlds, enabling efficient interaction between quantum and classical computing.

Quantum computing offers great potential. In a system where bits become qubits, the principles of quantum mechanics apply. These are the fundamental theories that govern the behaviour of the smallest known particles. This allows quantum systems to process information at unprecedented speed, solving complex problems far beyond the reach of even today’s most powerful supercomputers.

Current state of quantum computing

Quantum computers currently exist on a small scale, with chips containing tens to hundreds of qubits. Scaling up these systems is essential for reaching the quantum advantage but remains a significant challenge. To give an idea, IBM developed the IBM Eagle, which is a quantum chip running 127 qubits. The company predicts to reach 1,000+ qubits by 2027.

Unlike the gradual advancements in quantum hardware, quantum algorithms are evolving at a much faster pace. These algorithms are specifically designed for quantum systems but can also be simulated on classical computers— using quantum simulators running on powerful supercomputers. When successful, they will be able to solve certain problems significantly faster than their classical counterparts, even on classical hardware. This leads to hybrid quantum-classical systems, which leverage the strengths of both worlds.

The lack of compatibility between quantum and classical components is often the root cause of setbacks in quantum computing

Computer glue

Researcher Nishant Saurabh is working on ways to bridge the gap between these two worlds. He is contributing to quantum middleware development, crucial software that facilitates seamless communication and integration between the diverse components of a hybrid system. Some refer to middleware as the glue that holds together various system elements—such as databases, operating systems, and applications.

According to Saurabh, quantum middleware is indispensable for the success of hybrid systems, and with that the future of quantum computing. He points out that many quantum systems have struggled upon deployment, primarily because they were not compatible with existing classical systems and faced inherent quantum hardware limitations. "The lack of compatibility between quantum and classical components is often the root cause of these setbacks," he explains.

To unlock AI's full potential, we need quantum computing

Best of both worlds

Recently, Saurabh and his collaborators from the University of Munich, BMW, Princeton University and National Labs developed Pilot Quantum: middleware designed to manage resources—such as processors and memory—and workloads, including tasks and computations, across both quantum and classical systems in a unified manner. This middleware enables users to design and execute hybrid workflows, combining quantum and classical computing based on which is best suited for each specific task. The paper on this middleware was recently accepted at CCGrid2025, a prestigious conference in the field of computing research.

Technological revolution

According to Saurabh, it’s time to start thinking about the future. “Quantum developments are ongoing,” he says. “If we want to apply and scale these emerging technologies effectively, we must anticipate their evolution and find ways to integrate them. Only then can we achieve the quantum advantage.”

This quantum advantage, Saurabh believes, will cause a true technological revolution. “Take AI, for example. Right now, we’re utilizing only about five percent of the data AI could potentially process. To unlock its full potential, we need quantum computing.”