What was shown
The authors did not run anything. They worked out, on paper, what a future quantum computer would need to simulate molecules found in lithium-ion battery electrolytes. They picked the smallest case they judge too hard for classical computers. The central idea is the logical qubit: one dependable unit of calculation, built from many fragile physical parts that check each other, much like several clerks double-checking a single ledger entry. The paper counts these units and the T gates, the costly steps of the calculation. It then converts the count into the size of the machine and the running time.
“For the smallest instance that is classically intractable (full configuration interaction simulation of PF6⁻ using the cc-pVDZ basis at 1 mHartree precision), we estimated a total of 16,382 logical qubits and 2.32×10¹² T gates (see Table 4); after compilation, this corresponds to an estimated footprint of ~24 million RSGs and a runtime of less than 1 day, in the moderate error rate case (see Table 7).”
— Section IV Discussion, in: Fault-tolerant resource estimate for quantum chemical simulations: Case study on Li-ion battery electrolyte molecules1
- How it was done: a calculation on paper, not run on any machine
- Qubits: The estimate calls for 16,382 logical qubits, error-protected units of calculation, for the smallest case the paper calls classically intractable.
- Error correction: A future machine that corrects its own errors while it runs, built from light; no existing device is assumed.
- Compared against classical computing: no comparison stated in the paper
Who it affects
This concerns battery makers and chemical firms that want to simulate how electrolyte molecules behave, for example when designing new battery materials. For someone in that field today, nothing changes: no machine exists that could run this simulation. Classical computing remains the working tool for such questions. The paper is a sizing exercise for planners, not a service anyone can buy.
Chemicals and materials: McKinsey & Company estimates $450B to $800B by 2035 — value at stake for the industry, not vendor revenue (Quantum Technology Monitor, April 2026, p. 22).
This figure is an estimate, not a measurement: McKinsey & Company is a consultancy with its own quantum advisory business. It covers a whole use case, not the result of this paper.
What is still missing
The estimate is a calculation for a fault-tolerant photonic machine that the paper does not demonstrate. It states no comparison against a named classical method, so the advantage itself is not shown. The smallest case needs 16,382 logical qubits and 2.32×10¹² T gates, which after compilation means a footprint of ~24 million RSGs, light-based building blocks of the machine, and a runtime of less than 1 day. The proposed speed-up from injecting several magic states at once has not been simulated in detail.
Against published roadmaps: the task needs 2.32 trillion gates and 16,382 logical qubits. No announcement we have on file reaches that many gates. The furthest any of them goes:
- UK Department for Science, Innovation and Technology2 (a government setting a target for its own national programme): 1 trillion gates for 2035.
An announcement is not a delivered machine, and this one would still fall short of what the task needs. Roadmaps count gates of every kind, so a machine of that size is necessary for this task but may not be sufficient.
Who is working on it
The paper (April 2021) lists its authors' affiliations as: PsiQuantum, Palo Alto; Mercedes-Benz Research and Development North America, Sunnyvale, CA 94085, USA. That is who signed this work — it is not a ranking of who is ahead.
Strengthens the case for fusion-based quantum computing on photonic hardware with interleaving, a design that builds the machine by joining small light-based pieces, and for surface code architectures running several magic state factories in parallel.
How this brief was made
This brief was written by software, not by a person: a language model read the full text of the paper, and code then checked every quotation, number and name in this brief against that text before it was published. The rules are set out under Method. Responsible for the rules and for this page: Georg Singer. The paper is peer-reviewed, published in Phys. Rev. Research 4, 023019 (2022). Georg Singer holds no shares in any company or institution named in this brief (as of October 2026).