Universal Quantum builds trapped-ion quantum computers designed from the outset to reach a million qubits. Trapped ions offer the best coherence and gate fidelity of the leading modalities, but the conventional approach uses lasers to address individual ions, which becomes an optical engineering problem that does not obviously scale past modest system sizes. The company’s response is to use electronic rather than optical control on silicon, and to connect modules with fast links so that scale comes from adding modules instead of enlarging one trap.

That architectural choice is the whole thesis: it trades some near-term demonstration performance for a path that can be manufactured with semiconductor processes. Progress in the field is now limited less by qubit quality than by whether an architecture can grow by orders of magnitude without collapsing under its own control complexity. Backed since 2020, Universal Quantum is a long-horizon AI infrastructure position where the outcome depends on engineering scalability rather than incremental benchmark records. A manufacturable path matters more than early benchmarks if the eventual market is industrial-scale computation.