Photonic Quantum Computing: Companies, Startups & Research

Microsoft, PsiQuantum, Xanadu Quantum Technologies and Quantum Computing Inc appear among the organizations active in photonic quantum computing. PsiQuantum is developing a fault-tolerant system based on silicon photonics and fusion operations. Xanadu works on integrated photonics and continuous-variable architectures, while Quantum Computing Inc pursues optical quantum systems. Quandela and ORCA Computing are also named in the field’s startup and company data.

PsiQuantum, Xanadu Quantum Technologies and ORCA Computing are named startups developing photonic systems. PsiQuantum focuses on fault-tolerant computing with silicon photonics and fusion operations. Xanadu works on integrated photonics and hybrid resource states, and its collaboration with ASML concerns lithography for photonic hardware. ORCA Computing uses quantum memories to store and synchronize optical operations. These are examples of active approaches, not a ranking.

Massachusetts Institute of Technology, University of Chicago and the U.S. National Institute of Standards and Technology are among the research organizations active in the field. MIT studies single-photon emitters and integrated optical components. University of Chicago works on spin-memory architectures and tree-encoded fusion. NIST studies photonic components including integrated single-photon detectors. University of Science and Technology of China also develops processors and boson-sampling experiments.

Funding data records $662,124,301 across 548 grants since 2020. These figures describe grants associated with the field, rather than a complete measure of private investment or company financing. The broader data also records 58 funding rounds, but that total should not be treated as photonic-only investment without checking the underlying transactions.

Max Planck Gesellschaft Zur Foerderung Der Wissenschaften E.V., Xanadu Quantum Technologies, PsiQuantum and Pasqal Netherlands B.V. each have 8 patents in the listed patent-holder data. Yeda Research and Development Co. has 7, while Fujitsu Limited and Quandela each have 6. These counts identify the named leading holders in the available patent data.

Patent activity declined 43% in the last 12 months versus the previous 12 months, with 237 patents in the 24-month window. By contrast, publications grew 11% year over year, reaching 1,078 in the last year. The different directions suggest that research output is increasing even as patent activity contracts; they do not establish the commercial value or technical maturity of individual systems.

In September 2026, Xanadu Quantum Technologies and ASML announced a lithography collaboration, while a University of Science and Technology of China paper proposed Gaussian boson sampling for image recognition. A publication in August 2026 examined filtering multiphoton events from a single-photon emitter, addressing photon purity. Other recent work explores optical nonlinearities, hybrid resource states and applications such as portfolio optimization.

The main approaches include continuous-variable architectures, single-photon linear optics, fusion-based architectures, boson sampling and integrated photonic circuits. Continuous-variable systems encode information in properties of light, while single-photon systems manipulate individual photons. Fusion approaches use entangling measurements on photonic resource states. Boson sampling focuses on sampling tasks, and integrated circuits describe a hardware implementation that can support different architectures. Their trade-offs include loss, resource overhead and generality.

Application areas include quantum chemistry and materials, combinatorial optimization, financial modeling, quantum machine learning and quantum networks. Recent publications also explore image recognition with Gaussian boson sampling and portfolio optimization on photonic quantum computers. These areas are under investigation; the available application data does not establish that photonic systems outperform classical methods on practical workloads.