While everyone argues about whose supercooled chip has the fewest errors, a University of Tokyo spinout just got a telecoms giant to bankroll a very different bet: a quantum computer that runs at room temperature.
On Monday 3 August, NTT (Japan’s telecoms and research behemoth) signed a capital and business alliance with OptQC, a startup building an optical, or photonic, quantum computer. In plain terms, OptQC encodes quantum information in particles of light rather than in frozen superconducting circuits. The pitch is a fault-tolerant machine of the one-million-qubit class by fiscal 2030. The deal expands a research tie-up the pair first signed in November 2025, and this time NTT is putting equity in.
Why light instead of cold metal
Most of the headline quantum machines from the big American labs use superconducting qubits that only behave inside dilution refrigerators chilled to a hair above absolute zero, colder than deep space. That cooling is expensive, bulky and a nightmare to scale. OptQC’s photonic approach works at room temperature and normal atmospheric pressure, because photons do not need to be frozen into submission. If it holds up at scale, that deletes one of the ugliest cost lines in the entire industry.
A quick decode. A qubit is the quantum version of a bit, able to hold a blend of 0 and 1 at once. Fault-tolerant means the machine can correct its own errors faster than noise wrecks the sum, the threshold every serious player is chasing. A million of them, error-corrected, is roughly the scale theorists reckon you need for the genuinely world-changing jobs, like breaking today’s encryption or simulating new materials and drugs from first principles.
Read the timeline, not the headline
The alliance is real cash and real intent, but the million-qubit machine is a 2030 promise, and quantum timelines have a long history of slipping. The concrete near-term steps are more modest: a joint research project running through fiscal 2027 to lock down the architecture, co-creation trials with customer companies starting this fiscal year, and a 10,000-qubit-class proof of concept targeted for fiscal 2028. NTT is taking a strategic equity stake to bind the partnership across hardware design, supply chain and enterprise rollout.
What is worth clocking is who is buying. NTT is not a scrappy venture punt, it is a national-scale incumbent hedging its own future, and optical quantum computing sits neatly beside the optical-fibre and photonics expertise it already owns. When the plumbing giant invests in the plumbing, it is usually worth a second look.
Did you know: the photons used in optical quantum computing can carry quantum information down the same kind of glass fibre that already pipes your broadband, which is a big part of why telecoms firms keep circling the field.
Related on Top Tool Stack: PsiQuantum’s $125m DARPA deal · Google’s self-correcting quantum chip