Quantum computing’s promise has inspired a flurry of research into diverse hardware platforms, each vying to become the foundation of tomorrow’s superpowers. Erik Hosler, a photonics visionary driving PsiQuantum’s hardware roadmap, highlights the importance of choosing an architecture that not only performs but can be produced on a scale. His insights remind us that the ultimate test of any quantum technology is whether it can be manufactured efficiently and economically.
In the race to commercialize quantum computers, photonic and matter-based approaches have emerged as the leading contenders. Photonics leverages particles of light manipulated through silicon photonic circuits, while matter-based systems, such as superconducting qubits or trapped ions, use engineered atoms or circuits at cryogenic temperatures. Each path has unique strengths and manufacturing challenges. Understanding these differences is crucial for anticipating which technology will deliver real-world impact first.
The Strategic Divergence: Photonics vs Matter
Photonics-based quantum hardware encodes qubits in photons traveling through waveguides on silicon chips. This approach promises room-temperature operation in principle and taps into mature semiconductor fabrication techniques. By contrast, matter-based systems trap ions with electromagnetic fields or cool superconducting circuits to millikelvin temperatures. These platforms excel at qubit coherence and gate fidelity but demand specialized cryogenic equipment and bespoke fabrication processes.
Decision-makers must weigh these trade-offs through a manufacturing lens. Photonics can leverage high-volume silicon foundries, while matter-based devices often require custom materials, precision deposition, and specialized packaging. The question becomes: which platform aligns best with existing industrial ecosystems, and which can scale to millions of qubits without prohibitive cost?
Manufacturing Realities in Photonic Systems
Silicon photonics benefits from decades of semiconductor innovation. Standard processes, lithography, etching, and deposition can pattern photonic circuits with nanometer precision. The existing infrastructure accelerates prototyping and offers pathways to wafer-scale production. PsiQuantum aims to build a million-qubit system, with manufacturing already underway.
Erik Hosler shares, “PsiQuantum aims to build a million-qubit system, with manufacturing already underway.” It emphasizes that beginning production early allows feedback loops between design and fabrication, driving yield improvements. By integrating quantum photonic devices into conventional fabs, PsiQuantum and similar ventures hope to shrink the gap between lab-scale demonstrations and industrial output.
Challenges in Matter-Based Architectures
Matter-based qubit systems face a separate set of manufacturing hurdles. Superconducting qubits require ultra-pure aluminum or niobium films deposited on wafers, followed by precise patterning of Josephson junctions. Any contamination or microscopic defect can degrade coherence. Trapped-ion platforms demand microfabricated electrode arrays and vacuum-sealed chambers, with tight tolerances on electrode geometry to maintain stable ion traps.
Scaling these architectures beyond a few hundred qubits intensifies these challenges. Each additional qubit multiplies the complexity of wiring, control electronics, and cryogenic integration. While labs have built devices with over a hundred superconducting qubits, moving to the thousands or millions needed for error-corrected logic remains a formidable manufacturing endeavor.
Scaling and Infrastructure: On-chip and Off-chip
Whether photon- or matter-based qubit density on a single chip is limited by heat dissipation, control line routing, and signal crosstalk. Photonic circuits circumvent some of these issues by routing light rather than voltage pulses, reducing thermal load and electromagnetic interference. Off-chip components, lasers, detectors, and amplifiers can be modular and positioned outside the cryostat in a classical rack.
Matter-based systems often push more infrastructure inside the cryogenic environment. Superconducting qubits require hundreds of coaxial cables and attenuators, all passing through multiple temperature stages. Trapped ions include laser delivery optics and vacuum hardware within or adjacent to cryostats. Each additional qubit thus increases the demands on thermal budgets and mechanical stability.
Leveraging the Semiconductor Ecosystem
Photonics developers aim to ride the coattails of high-volume CMOS fabs. Foundries that produce millions of transistors per wafer can, with modest process adjustments, pattern silicon waveguides and modulators. This compatibility offers a clear manufacturing edge: existing supply chains, design tools, and quality controls can be reused rather than reinvented.
Matter-based qubit makers are forging new supply chains. Superconducting devices require vendor partnerships for ultra-high-purity materials and specialized deposition tools. Trapped ions depend on MEMS foundries capable of fabricating electrode arrays and micro-optics. Building these ecosystems takes time and investment, whereas photonics can more readily plug into an established industrial backbone.
Quality Control and Yield Considerations
High yields are essential for any large-scale manufacturing effort. Silicon photonic processes in leading fabs routinely achieve wafer yields above 90%. Quantum photonic chips must meet similarly stringent yield targets, with each defective waveguide or coupler representing a lost qubit. Early PsiQuantum production runs focus on mapping defect distributions and optimizing process recipes to boost functional device counts.
In matter-based workflows, yield challenges include film uniformity, junction reproducibility, and trap performance. Even small variations can lead to qubit frequency mismatches or increased decoherence. Quality control protocols ranging from in-fabrication metrology to post-fabrication cryogenic testing are still being refined for quantum hardware, and scaling them to thousands of units will be a major undertaking.
Early Ecosystem and Supply Chain Developments
Some ecosystem players are already emerging. Specialty foundries offer multi-project wafer runs for photonic quantum chips, allowing startups to evaluate designs without full-scale production commitments. Packaging firms are developing turnkey photonic modules that integrate lasers, detectors, and control photonics in a compact form factor.
On the matter side, consortia of academic labs and industrial partners are standardizing processes for superconducting junction fabrication and trap electrode arrays. Collaborative roadmaps outline shared tool requirements, materials specs, and metrology standards. Over time, these initiatives aim to lower barriers to entry and accelerate manufacturing scalability.
Bridging the Gap: Hybrid Approaches
Recognizing that no single platform may dominate, some companies explore hybrid quantum architectures. Photonic interconnects could link superconducting or trapped-ion modules, combining the best of both worlds: high-fidelity gates on matter qubits with scalable routing via light.
Such hybrid systems present their manufacturing puzzles, integrating disparate materials and aligning optical and electrical subsystems, but they may offer a practical route to large-scale quantum networks.
Manufacturing as the True Differentiator
Quantum computing’s future hinges not only on qubit performance but also on the ability to produce devices at scale. While matter-based systems have demonstrated impressive coherence and gate fidelity, photonic platforms stand to benefit from existing semiconductor infrastructure, high-yield processes, and modular off-chip components.
As PsiQuantum and other pioneers push toward million-qubit assemblies, the manufacturing edge will determine which architecture first crosses the chasm from lab curiosity to commercial imperative. The next quantum leap may not come from physics alone but from mastering the industrial art of large-scale production.
