Software, Not Qubits, Will Drive Quantum Computing’s Next Breakthrough, Horizon Quantum CEO Argues
Horizon Quantum Computing Pte. (NASDAQ:HQ) is positioning itself as a software-centric quantum company, betting that a hardware-agnostic development environment—not more qubits—will be the key to unlocking practical quantum advantage. During a Canaccord Genuity event, Chairman and CEO Dr. Joseph Fitzsimons argued that the industry has fixated too long on hardware milestones while neglecting the software stack needed to turn raw quantum capability into real-world solutions.
The conversation, hosted by Canaccord analyst Kingsley Crane, came at a moment of heightened interest in quantum computing. Crane noted that the U.S. Department of Commerce recently announced $2 billion in awards to nine quantum companies, and that sovereign capital is “flowing in at scale.” He also pointed to recent breakthroughs in quantum error correction, which have sharpened the debate over when—and how—quantum systems will deliver on their promise.
Crane described Horizon Quantum as “more of a pure software play” in a field dominated by hardware builders. The company, which went public in March after merging with dMY Squared, trades on Nasdaq under the ticker HQ.
Fitzsimons drew a parallel with the early days of classical computing, noting that hardware alone is useless without software to direct it. “It’s not enough to simply have a number of well-controlled qubits,” he said. “You actually need to decide what you want to do with them to turn that to solve a particular problem.”
Horizon’s goal is to create a path toward the “automatic acceleration” of classical code—letting domain experts in finance, pharmaceuticals, machine learning, energy, aerospace, and defense harness quantum power without becoming quantum specialists themselves. The company is building tools that automate the construction of quantum algorithms from conventional programming languages.
Fitzsimons acknowledged that the software side of quantum computing has historically been underprioritized. “Yes, absolutely,” he said when asked. Hardware has attracted more attention because it involves exotic physics and specialized systems that classical computers cannot easily simulate. But Fitzsimons argues that the software challenge is equally significant, and that Horizon is essentially trying to compress 80 years of classical computing software evolution into eight years.
To that end, the company’s science team is twice as large as its engineering team, reflecting the deep research still needed on the software side. Fitzsimons described Horizon’s stack as broader than existing quantum programming frameworks, which have typically evolved by exposing new hardware capabilities piecemeal. Horizon is taking the opposite approach: “We’re saying, how does a perfect quantum computer work? How do we program that ideal system? How do we get that code to run on today’s computers?”
The company’s Triple Alpha platform, named after the first three letters of the Greek alphabet, is designed to support capabilities that many current-generation systems lack—including general control flow, indefinite loops, branching programs with indefinite runtime, mid-circuit measurement, and classical computation based on those measurements. When today’s hardware cannot directly support those features, Horizon stitches together multiple hardware runs through classical software to simulate the effect of a more advanced quantum computer. That approach increases the number of hardware calls but gives existing machines capabilities usually associated with more mature systems.
Fitzsimons also highlighted Horizon’s software roadmap: Hydrogen, a portable assembly language; Helium, an imperative language; and Beryllium, an object-oriented language previewed in December. Beryllium lets developers mix classical and quantum data and define how information is represented and manipulated. A developer could build a library for a data type like a matrix or tensor, and another developer could import that library without needing to understand quantum mechanics underneath.
That library model, Fitzsimons said, could make quantum programming more accessible and create network effects similar to Python’s ecosystem in classical computing. He noted that it also insulates developers from hardware technology risk: “They don’t need to pick which hardware system is going to win.”
Although Horizon is not primarily a hardware company, it has invested in physical systems to support its software work. The company operates a superconducting system in Singapore built from components, currently using a Rigetti processor and a Quantum Machines controller. Crane noted that Horizon recently purchased one of IonQ’s 256-qubit systems for a multimodality test bed. Fitzsimons said that system gives the company exposure to trapped-ion technology, which involves different operational and control requirements—including cold systems requiring microwave control and systems requiring high vacuum and optical stability. He added that if error rates improve sufficiently, the system could fall in the range where real quantum advantage is possible, though he cautioned that Horizon does not yet know whether it will achieve that.
In closing, Fitzsimons emphasized the economic difference between Horizon’s software business model and that of hardware-focused quantum companies. By allowing developers to write code in one language and run it on multiple machines, Horizon aims to insulate its customers from hardware volatility and accelerate the timeline for practical quantum adoption.
This story is based on a presentation at a Canaccord Genuity event and has been edited for length and clarity.