The NISQ Era
NISQ describes today's noisy, intermediate-scale quantum processors: tens to hundreds of qubits without full error correction.
What NISQ means
John Preskill coined Noisy Intermediate-Scale Quantum (NISQ) in 2018 to name the present phase of quantum computing. Intermediate-scale means roughly 50 to a few hundred qubits, enough that classical simulation becomes difficult but far short of the millions expected for large error-corrected machines. Noisy means gate and readout errors are too high to run deep circuits reliably, and there are too few qubits for full error correction.
The central constraint
Without error correction, errors accumulate with every gate, so only shallow circuits give trustworthy results. The usable circuit depth is set by the ratio of coherence time to gate time and by two-qubit gate fidelity. This bounds which algorithms are feasible now and rules out textbook algorithms like Shor's at useful sizes.
NISQ-era approaches
- Variational algorithms (VQE, QAOA) that use shallow parameterized circuits with a classical optimizer
- Quantum simulation of physics and chemistry, the most natural early fit
- Error mitigation, which reduces bias in results without full correction
- Sampling benchmarks that demonstrate quantum advantage on contrived tasks
Error mitigation versus correction
Error mitigation (zero-noise extrapolation, probabilistic error cancellation) improves expectation values by running extra circuits and post-processing, but its overhead grows fast with circuit size, so it extends NISQ rather than transcending it. True fault tolerance needs error correction and many more qubits.
NISQ is a transition, not a destination. Its value is scientific exploration and hardware maturation; the field's stated goal is to leave it for fault-tolerant machines, and Kronos treats present quantum results as one input cross-checked against classical simulation, not a standalone authority.