PsiQuantum is a Photonic Qubits company founded in 2016 and based in Palo Alto, United States. It has raised $2.3B in total funding, most recently a Series E in 2025 at a $7B valuation.
| Date | Stage | Amount | Valuation | Lead investors |
|---|---|---|---|---|
| Sep 10, 2025 | Series E | $1B | $7B | BlackRock, Temasek, Baillie Gifford |
| Apr 30, 2024 | Growth | $620M | — |
| Australian Government, Queensland Government |
| Jul 27, 2021 | Series D | $450M | — | BlackRock |
| Jan 1, 2020 | Series C | $150M | — | Atomico, BlackRock |
| Jan 1, 2017 | Series B | $50M | — | — |
| Jan 1, 2016 | Series A | $13.5M | — | Playground Global |

PsiQuantum demonstrates an 8x8 optical switch with 1.2 dB insertion loss aimed at AI data center networking.

PsiQuantum's 8x8 lab prototype optical switch for AI datacenter GPU networks loses less light than rivals, with larger devices planned.
PsiQuantum demonstrates an 8×8 optical circuit switch with 1.2 dB typical loss for AI datacentre networks.

PsiQuantum unveils an 8×8 optical circuit switch demonstration vehicle built on its silicon photonics quantum architecture.

PsiQuantum unveils a low-loss silicon photonics optical switch for its photonic quantum computing stack.

PsiQuantum unveils an 8×8 optical circuit switch for AI data centers, with 1.2 dB insertion loss and under 3 ms reconfiguration in lab tests.
Silicon photonics wafer fabricated using standard CMOS processes at GlobalFoundries, announced February 2025. Designed as the manufacturing-ready hardware platform for PsiQuantum's fault-tolerant photonic quantum computer, integrating single-photon sources, detectors, and routing circuits on a single chip manufacturable at semiconductor scale. Serves as the physical substrate for PsiQuantum's surface-code error correction architecture targeting one million physical qubits. Core to the Chicago and Brisbane facility buildouts announced in 2026.
Software development platform for fault-tolerant quantum algorithm design, launched September 2025. Supports the full algorithm lifecycle: high-level specification, automatic mapping onto surface-code error correction, qubit routing optimization across the Omega photonic chip, and resource estimation (logical qubit count, circuit depth). Integrates NVIDIA CUDA-Q for 450x simulation speedup. The only software platform purpose-built for fault-tolerant algorithm development; targeted at researchers and enterprises preparing algorithms for near-term fault-tolerant hardware.