Feynman's Case for Quantum Simulation
Richard Feynman's 1981 argument that classical machines cannot efficiently simulate quantum systems, and that quantum machines can.
The motivation
In his 1981 lecture Simulating Physics with Computers, Richard Feynman observed that the state of a quantum system with n particles lives in a Hilbert space whose dimension grows as 2^n (for n two-level systems). Storing an arbitrary state vector therefore requires exponentially many classical numbers, and evolving it means multiplying by exponentially large matrices.
Feynman's conclusion was blunt: nature is not classical, and if you want to simulate it, you had better make the simulator quantum mechanical. A computer built from quantum degrees of freedom would represent a 2^n-dimensional state using only n physical qubits, sidestepping the exponential storage wall.
Why the exponential appears
The trouble is entanglement and superposition. A separable product state of n qubits can be described by O(n) parameters, but a generic entangled state cannot be factored, so its amplitudes must be tracked jointly. Interference between amplitudes is exactly what classical sampling struggles to reproduce, because probabilities can cancel, not merely add.
What a quantum simulator provides
A controllable quantum system whose interactions can be programmed to match a target Hamiltonian evolves under the same laws you wish to study. Measurements then estimate observables. The simulator does not compute the full state vector explicitly; it lets physics do the linear algebra, and you sample the answers you need.
- Digital approach: universal gate sets approximate arbitrary evolution.
- Analog approach: engineer a device whose native dynamics mirror the target.
- Both trade an intractable classical cost for a tractable quantum experiment.
This idea seeded the field. It reframed the goal of quantum computing from abstract logic toward a concrete physical payoff: chemistry, materials, condensed matter, and eventually the many-body kinetic systems that appear in fusion plasmas. Kronos treats classical simulation as the workhorse today; quantum simulation is a research horizon for problems where the state space defeats classical hardware.