Guest post by Zeynep Koruturk, Dr. Kris Naudts, & Donald Harmitt of Firgun Ventures and Professor Bob Coecke of Relational Intelligence Limited Few questions in the quantum sector are asked more often, or answered more loosely, than how many qubits a useful quantum computer will require. Press releases announce systems with hundreds of qubits, then thousands, and the implicit suggestion is that the field is approaching a finish line. The reality is more nuanced, and arguably more interesting. It is not as black and white as saying there is a single number of qubits that makes a quantum computer “useful”. The right number depends entirely on what you want the machine to do, on which type of machine you are using, and on how cleverly the underlying problem has been formulated. Not all quantum bits (or qubits) are created equal. A qubit is the basic building block of a quantum computer, the equivalent of a bit in a classical machine, but qubits are fragile and error prone. To do any serious computation, many physical qubits must be bundled together to create one reliable “logical qubit” through a process called quantum error correction. The ratio between the two, depending on the technology and the target error rate, can range from roughly 2 to 1 at the optimistic end to over 2,000 to 1 at the demanding end. Rather than asking how many qubits quantum computing needs in the abstract, we instead reframe the question to how many qubits each specific application needs on each specific type of machine and then organise the answers into three broad bands of ambition and difficulty. Why Qubit Numbers Sometimes Mislead A small chemistry calculation might need only a few dozen highly reliable qubits. Breaking modern encryption or proving advantage on hard optimisation problems (finding the