Quantum Simulation

Molecules are quantum, and simulating them classically blows up exponentially — so use a quantum computer, which speaks their native language.

Quantum Computing · Lesson 11 · 10 min read

Designing a new drug or a better battery means predicting how molecules behave. But molecules are quantum, and simulating even a modest one exactly would overwhelm every supercomputer on Earth — the math explodes exponentially with each electron. So chemists mostly guess and test in the lab, slowly and expensively. Back in 1981 Feynman pointed at the obvious fix: if nature is quantum, build a quantum computer to imitate it. Hold the question: why is simulating molecules impossibly hard for classical machines, and why would a quantum computer find it natural?

Why molecules break classical computers

Describing a quantum system of n interacting particles takes roughly 2ⁿ numbers — the same exponential blow-up as n qubits (L04). That’s because entangled electrons can’t be described one at a time; you need the whole joint state. Ten electrons is about a thousand numbers (easy). Fifty is about 10¹⁵ (a supercomputer strains). A hundred is more numbers than there are atoms on Earth. Real molecules blow past that fast, so exact classical simulation is hopeless and chemists fall back on approximations that often fail exactly where it matters.

Use quantum to simulate quantum (Feynman's idea)

A quantum computer’s qubits already live in superposition and entanglement — the very stuff a molecule’s electrons are made of. So instead of storing 2ⁿ numbers, you map the molecule’s state onto ~n qubits and let them evolve under the same rules. The simulator speaks the system’s native language: representing a state that needs 2ⁿ classical numbers takes only about n qubits. The exponential wall that stops classical chemistry is simply the space a quantum computer lives in for free.

What it unlocks — and the honest timeline

The payoffs are concrete: better catalysts (cheaper fertilizer means cheaper food), improved batteries and materials, understanding how a drug molecule binds, and cracking puzzles like high-temperature superconductors. Many people think this — not code-breaking — is quantum computing’s biggest eventual prize, and possibly an earlier one, since useful chemistry may need fewer and less perfect qubits than breaking RSA. The honest caveat: today’s machines do small proof-of-concept simulations, not industrial chemistry yet. The direction is real; the scale isn’t here.

Watch the classical blow-up, then the quantum fix:
1. One electron’s relevant state: a handful of numbers — easy.
2. Add electrons and the possibilities multiply rather than add, because entanglement forbids treating them separately: n electrons need on the order of 2ⁿ numbers.
3. 10 electrons → ~1,000 numbers (fine). 50 → ~10¹⁵ (a supercomputer struggles). 100 → ~10³⁰, more than all classical memory on Earth. A single caffeine molecule has far more than 100 relevant electrons.
4. Quantum fix: map the degrees of freedom onto qubits — roughly n qubits hold what needed 2ⁿ classical numbers (L04 in reverse). About 100 good qubits could represent a state no classical computer ever could.
5. So the exponential wall that halts classical chemistry is exactly the exponential room a quantum computer occupies naturally — the match Feynman saw.