Translating technical progress into investor-relevant judgment, one note at a time.
IonQ reaches breakeven with codes that pack four logical qubits into an eighteen-qubit block, on a chain that measures itself without moving an ion. The larger codes did not do better, and the machinery of correction is now a major error source in its own right.
Quantinuum’s thermometer misses its own prediction by more than six standard deviations, and Pasqal’s model meets an offset it never contained. Both teams went looking for the miss instead of tuning it away.
A two-qubit fidelity tells you how often a machine errs, not what kind of error it makes. Two papers show the same hardware limited by different imperfections depending on the job it is given.
IBM Research and the University of Chicago build a fidelity certificate into a classically hard circuit, reading it from the code’s own error checks rather than from a model of the hardware. It costs almost every shot.
Qedma, IBM, RIKEN and BlueQubit drive a magnet on 51 and 74 qubits. The classically simulable sizes cannot say whether its rhythm survives at scale; the two error-mitigated points can.
Quantum computing will transform certain industries, but the path is long, uneven, and filled with overhyped milestones. Our edge comes from deep technical fluency, ruthless skepticism, and a focus on durable business value.
Tracking the frontier of quantum hardware, algorithms, and error correction.
Assessing company execution, roadmaps, team, and commercial traction.
Synthesizing signal into durable insights for long-term capital allocation.
Open the notes index and pick a “start here” note, or any question that pulls you in.
Every note links to the notes it leans on; the pink links weave the web.
The Research Board connects a question to findings, limitations, and the sources behind them.
The weekly papers track the frontier one publication at a time; the company files read the businesses behind the tickers.
Take a random plunge into the notebook, or browse by category.
Weekly Publications · 12–18 September 2026
A weekly perspective on the technologies attracting research attention and what they could mean for public companies.
Counts from 148 reviewed publications · rounded shares in brackets · categories overlap
