PsiQuantum

PsiQuantum is a quantum-computing company pursuing photonic quantum computing using silicon-photonics manufacturing. Founded around 2016, it states a goal of building a large-scale, fault-tolerant machine with roughly one million physical qubits, and works with the semiconductor foundry GlobalFoundries to fabricate its chips. Its stated targets are goals, not demonstrated results.

#Photonic and measurement-based approach

PsiQuantum encodes quantum information in single photons rather than in superconducting circuits or trapped ions. Photons interact weakly, which makes deterministic two-qubit gates hard, so the company follows a measurement-based strategy in which computation proceeds by preparing entangled resource states and performing measurements. Its published work develops fusion-based quantum computation, a variant in which small entangled states are stitched together by probabilistic "fusion" measurements to build up a fault-tolerant computation (Bartolucci et al. 2021). This approach is designed around the strengths and weaknesses of photonics: photons can be routed and networked and operate without millikelvin cryogenics for the optical components, but photon loss and the probabilistic nature of the gates are the central challenges.

#Silicon-photonics manufacturing

PsiQuantum's strategy emphasizes manufacturability. Rather than building a small demonstration processor first, the company has focused on fabricating photonic components on standard silicon wafers at GlobalFoundries, arguing that a fault-tolerant machine will need billions of components produced at semiconductor-industry volumes. The bet is that leveraging an existing high-volume foundry process is the credible path to the scale that error correction requires. The architecture still needs cryogenic cooling for its single-photon detectors, though far less demanding than the millikelvin environment superconducting qubits require. The company has described plans for large facilities, including sites associated with Brisbane in Australia and Chicago in the United States, framed as future utility-scale installations rather than operating machines. As of early 2026, this manufacturing-first approach means the company has published less about intermediate small-qubit processors than some competitors, and its headline million-qubit figure describes an objective rather than an achieved system.

#Fault-tolerance goal

Because single photons are lossy and photonic gates are probabilistic, PsiQuantum's architecture is built around heavy Quantum error correction from the outset. The company argues that useful applications require fault tolerance and therefore aims directly at a large error-corrected machine rather than at near-term noisy demonstrations. This distinguishes its public positioning from companies that emphasize present-day noisy intermediate-scale devices. Whether the manufacturing-first, fault-tolerance-first path reaches a working machine on the company's stated timeline is unproven; it is a goal that depends on unsolved engineering, and it should be read as such.

#Relation to cryptography

A fault-tolerant photonic machine of the scale PsiQuantum targets could in principle run Shor's algorithm, which is the connection to the Post-quantum cryptography threat model. No such machine exists as of early 2026, from PsiQuantum or anyone else. The company's relevance to cryptography is entirely prospective and contingent on its long-range goals being met, which remains uncertain.