What happened
- A group of researchers from the Open Quantum Institute, hosted at CERN, Google Quantum AI, Microsoft, Oratomic, Nvidia, JPMorganChase, MIT and Harvard, among other institutions, put together a guide to 24 problems with credible prospects for quantum advantage. The Quantum Insider reported it on 6 November 2026.
- For each problem they describe the mathematical structure, the data required and the conditions that preserve the quantum machine’s edge.
- The work, posted on the preprint server arXiv, is a synthesis of existing research and not a new experiment.
- The problems span number theory, optimization, statistical inference, quantum simulation, linear algebra and the analysis of shapes and connections in data.
Why it matters
- A theoretical speed advantage does not guarantee faster results in practice. A quantum algorithm may need far fewer steps than a classical one and still take longer to give a useful answer.
- Preparing data, correcting errors and reading results can erase the savings, so the whole calculation has to be assessed against the best classical methods.
- The guide helps scientists and companies locate where the theoretical gain could become useful, instead of looking for a quantum solution to any problem.
The number
15 hours and 74 million physical qubits. For one specific random constraint-satisfaction problem, the paper cites an analysis placing the quantum-classical crossover at about 15 hours, with some 74 million physical qubits, under a given error rate and correction design. It shows the distance between a mathematical advantage and an affordable computation.
Context
The authors concentrate on speedups greater than quadratic, a threshold they consider probably needed to offset the high costs of running early quantum computers with error correction. A quadratic speedup means that if a classical method needs a million steps, a quantum one might need about a thousand, but that does not imply finishing a thousand times sooner, because each quantum step can cost more.
The guide distinguishes types of evidence: rigorous mathematical results, evidence limited to certain classes of classical methods and numerical or heuristic studies. The comparisons are, in general, against the best classical methods known today, which can also improve. In linear algebra, for instance, the exponential improvement requires that the numbers can be loaded quickly and that the requested output be a general measure rather than every value.
The 24 problems are grouped in two stages according to how much is known about their potential. The team’s recommendation is to start from tasks where the advantage already has support and then look for real problems that fit, rather than starting from an urgent challenge and searching for a quantum solution.
At Mamífero we follow quantum computing. We reported that a trapped ion corrected its own errors and passed break-even and that IBM, Oak Ridge and AMD join quantum and classical computing to correct errors. Also that Telefónica and IQM will install Madrid’s first quantum computer and that Vietnam will get its first quantum computer with a photonic machine.
What’s next
- The researchers hope the guide will steer the search for applications where the advantage already has support.
- Preparing initial states and defining which systems are hard for classical methods appear as open challenges.
- Classical methods will keep improving, which can shrink or erase some advantages.
Bottom line
The quantum promise is not that everything speeds up, but that a few problems with the right structure get solved far faster. This guide puts a name on 24 of them and reminds us what each one costs.
Sources
- The Quantum Insider, “Study Maps 24 Problems Where Quantum Computers Could Gain an Edge”, 6 November 2026, link. It cites an arXiv preprint without peer review.
- Cover image: Anita Fors (Chalmers), CC BY-SA 4.0, via Wikimedia Commons
Edited by Rodrigo Cornejo. How we select and verify: who writes these notes.
