ORCA Computing
UK photonic quantum-computing company spun out of the University of Oxford in 2019. Distinguishing technology is a room-temperature photonic quantum memory based on Raman storage in warm atomic vapour, originally developed in the Walmsley group at Oxford. The memory provides controllable, low-loss interactions between travelling photons and atomic ensembles, enabling synchronisation and "repeat-until-success" operations in a photonic processor. Ships rack-mounted photonic appliances (PT-1, PT-2) targeting quantum machine learning and sampling workloads, integrated with classical GPU clusters.
Timeline
Current flagship
ORCA PT-2 — photonic processor with integrated quantum memory; rack-mounted
Milestones
- 2022-20Nine PT-1 photonic systems deployed to UK Ministry of Defence, the UK NQCC, Poznan Supercomputing Center, and other academic / government sites — first commercial deployments of an ORCA photonic appliance
- 2025-12Demonstrates use of PT-series photonic processor as a quantum random-number / entropy source for post-quantum cybersecurity workflows
- 2025-11USD 37M Series B closed with Temasek, Baillie Gifford, NVIDIA's NVentures, the Qatar Investment Authority, and Morgan Stanley
- 2025ORCA delivers PT-2 photonic system to the UK National Quantum Computing Centre (NQCC) Photonic Testbed; demonstrates application to Vodafone fibre-network route optimisation
- 2024Collaboration with Poznan Supercomputing and Networking Center (PCSS) on a quantum-AI data-centre integration blueprint built on NVIDIA infrastructure
- 2019Founded as Oxford spin-out by Richard Murray (CEO), Cristina Escoda, Ian Walmsley, and Josh Nunn
Roadmap
- 2026 PT-3 commercial photonic system, reported to match ~180 GPUs of throughput on certain quantum machine-learning workloads source ↗
Capability details
Quantum computing
- Qubit type
- photonic
- Physical qubits
- —
- Logical qubits
- 0
- 1Q gate fidelity
- —
- 2Q gate fidelity
- —
- T₁
- —
- T₂
- —
- EC code
- —
- Connectivity
- —
Quantum memory
- Platform
- warm-atomic-vapour (Raman / GEM)
- Storage time
- —
- Retrieval efficiency
- —
- Fidelity
- —
- Wavelength
- —
- Mode capacity
- —
14 unverified fields
-
modalities.qc.physical_qubits_current -
modalities.qc.one_q_fidelity -
modalities.qc.two_q_fidelity -
modalities.qc.connectivity -
modalities.memory.storage_time_ms -
modalities.memory.retrieval_efficiency -
modalities.memory.fidelity -
modalities.memory.wavelength_nm -
modalities.memory.mode_capacity -
positioning— The user's brief referenced "boson-sampling / squeezed light" — ORCA's public materials emphasise a memory-based photonic architecture and single-photon/coherent-light operation rather than squeezed-light Gaussian boson sampling per se. Architectural detail beyond the Raman/GEM memory not verified against a peer-reviewed source in this run. -
shareholders -
headcount -
revenue -
milestones[NSCC Singapore]— Brief mentioned NSCC Singapore deployment; not found in search results. Omitted rather than fabricated.
People
References
- web https://orcacomputing.com/orca-pt-2/ PT-2 product page
- web https://orcacomputing.com/orca-pt-1/ PT-1 product page and deployment history
- press https://thequantuminsider.com/2026/03/24/11-companies-lighting-up-the-quantum-photonics-sector/ Photonic-sector overview citing the Nov 2025 Series B and PT-3 roadmap
- press https://www.hpcwire.com/2025/12/10/orca-turns-its-quantum-photonic-machine-to-cybersecurity/ PT-series entropy / cybersecurity application (Dec 2025)
- press https://thequantuminsider.com/2021/04/08/tqd-exclusive-interview-with-richard-murray-ceo-of-orca-computing/ CEO interview covering memory-based photonic architecture rationale