Quantum computing has been the subject of both excitement and skepticism for years. Advocates tout its promise as the first new computing architecture in 80 years, while critics argue that quantum’s considerable technical challenges will keep the technology out of the mainstream for decades. Q-CTRL Pty. Ltd. is trying to shift that conversation from theory to utility.

The Los Angeles- and Sydney-based quantum infrastructure software company this week announced what it calls the first demonstration of “practical quantum advantage” using publicly available hardware from IBM Corp. Working on a commercially practical materials science problem involving electron behavior in advanced materials, Q-CTRL said it achieved a 3,000-fold performance advantage over the best conventional computing alternative while maintaining acceptable accuracy.

For Chief Executive Michael Biercuk (pictured), the significance of the announcement extends beyond benchmarks. He believes the industry has already crossed the chasm of making quantum systems useful for solving real-world problems, particularly in areas such as chemistry, materials science, navigation and optimization.

“The practical machines are already here,” Biercuk said in an interview at IBM’s Think 2026 conference in Boston. “We made an IBM machine better than the best conventional alternative for a real problem that people care about.”

Quantum advantage

The latest demonstration involved simulations of materials in which electrons interact strongly with one another, a notoriously difficult class of problems for conventional supercomputers. Understanding them can advance research into areas such as superconductivity, high-density battery chemistry and advanced photovoltaic behavior.

One of the most tantalizing mysteries is the behavior of high-temperature superconductors, which conduct electricity without resistance at relatively warm temperatures, yet whose underlying physics remains poorly understood.

“We just don’t know why,” Biercuk said of the materials first discovered in the 1980s.

Classical simulations become computationally overwhelming because electron interactions scale exponentially with system size. Quantum computers, however, obey the same quantum mechanical rules as the materials themselves, making them better suited for calculating interactions at scale.

Biercuk’s optimism runs counter to the common narrative that quantum computing remains years away from practical use. Many researchers still point to persistent technical obstacles, including the fragility of the fundamental unit of quantum information called qubits, high error rates and the extreme cooling requirements of quantum hardware.

Solving those problems may take years – if they can be solved at all – but Biercuk argues that software can overcome many of the thorniest issues today. He drew an analogy to error-correcting algorithms that compensate for data corruption caused by noise, interference or physical defects in semiconductors.

Founded nearly nine years ago by Biercuk, a Harvard physical Ph.D. and former professor specializing in quantum control engineering, Q-CTRL company develops infrastructure software designed to stabilize and optimize quantum systems. Rather than building its own quantum computers, Q-CTRL focuses on improving the performance of existing hardware platforms.

Making hardware sing

“The core to it is the infrastructure software that we layer on top of the platform,” Biercuk said. “Software is what makes the hardware sing.”

Q-CTRL’s software suppresses errors and optimizes the use of qubits, allowing existing quantum hardware to execute far larger and more accurate calculations than previously possible. Its software pipeline automatically handles tasks such as selecting the best qubits for a given algorithm, reducing interference between qubits and minimizing measurement errors.

The company says those optimizations enable it to run algorithms involving more than 14,000 entangling operations, in which particles share a single far quantum state, causing the actions of one to affect others instantly. Entanglement is what enables quantum computers to achieve exponential processing power.

Q-CTRL has demonstrated commercial viability in other areas. Last year, it unveiled a GPS-independent navigation system that uses quantum sensors and software-based error suppression to detect subtle variations in Earth’s magnetic field. The technology can be used as a backup navigation aid when GPS signals are unavailable or jammed.

The company is now shifting its focus from proving accuracy to exploring new scientific territory.

“Now that we know it’s accurate to within 1% of the best alternative tool, and we’re able to show in certain regimes the physics that we know is true, we now get to explore the unknown,” Biercuk said.

He pointed to high-energy-density battery materials, photovoltaic systems and chemical dynamics as areas where quantum simulations could accelerate discovery. For example, researchers may eventually be able to model how light interacts with exotic materials or how new compounds behave before they are physically synthesized. That could enable the virtual discovery of new compounds before physical synthesis, accelerating research cycles from years to months while drastically reducing costs.

Quantum computing is also useful in optimization problems such as logistics routing, transportation scheduling and military convoy planning.

Q-CTRL’s technology has attracted customers, including Lockheed Martin Corp. and Airbus SAS, with early commercial deployments already in the field.

Software compensation

Like the materials science project, the navigation system relies heavily on software to compensate for hardware imperfections. Biercuk views that as a recurring theme across the industry. “We knocked down those walls with our software,” he said.

Biercuk doesn’t believe quantum computers will become general-purpose replacements for conventional processors. Instead, he sees them evolving into specialized accelerators integrated into hybrid workflows, much like graphics processors complement CPUs today.

“There’s a long list of things quantum isn’t good at,” he said. “We don’t think of it as a generalist processor any more than you run Windows on a GPU.” The key innovation will be high-level languages that allow developers to transparently intermix operations on quantum and conventional hardware.

“The way you manipulate the machines today is using the equivalent of assembly language,” he said. “It’s your responsibility to move data around, choose operations, schedule operations and reduce errors. We believe the key is abstraction, so that people with generalist IT skills can take advantage of quantum computers in their existing workplace.”

Q-CTRL’s latest results are likely to intensify debate over how close quantum computing truly is to commercial relevance. People attending the Think conference heard IBM Chief Executive Arvind Krishna make a forceful case that the time is now.

“The people who dismiss it think of it as being a science problem that’s not going to get solved,” he said in his Tuesday keynote. “That’s no longer true. It’s an engineering problem today. And when something moves from science to engineering, it is not a question of whether it will be there. We believe quantum advantage will be reached this year.”

Photo: Paul Gillin/SiliconANGLE