Simulating the Quantum World

Chemists and physicists want to predict how a molecule will behave: how it bonds, reacts, absorbs light. In principle the rules are known — quantum mechanics — so why not just compute the answer on a powerful computer?

Here's the wall. To describe n quantum particles, you don't track n numbers; you track the amplitudes of every possible configuration at once — and there are $2^{n}$ of them. Ten particles need about a thousand numbers. Fifty need more than a quadrillion. A few hundred, and the description has more numbers than there are atoms in the universe. The information simply won't fit in any classical machine.

This is quantum simulation, and its difficulty is famous: in 1981 Richard Feynman pointed at exactly this explosion and argued that to simulate nature you'd need a computer that is quantum. That observation launched the entire field of quantum computing.

Try it. Slide the number of quantum particles (qubits) and watch the state size — the count of amplitudes you'd have to store — and the memory it would need. For a few particles, see the actual state vector drawn as bars.

<p class="hint">{{hint}}</p>
<label class="slab">{{label_particles}} <b id="nval">3</b>
  <input id="n" type="range" min="1" max="300" value="3" />
</label>
<div class="stats">
  <div class="stat"><span>{{stat_amplitudes}}</span><b id="amps">8</b></div>
  <div class="stat"><span>{{stat_memory}}</span><b id="mem">128 bytes</b></div>
</div>
<div id="cmp" class="cmp"></div>
<div id="bars" class="bars"></div>

* { box-sizing: border-box; }
body { font-family: system-ui, sans-serif; color: #222; margin: 0; }
.hint { font-size: .9rem; color: #444; margin: 0 0 .7rem; line-height: 1.45; }
.slab { display: block; font: 600 14px system-ui; color: #1d3557; margin: .4rem 0; }
.slab input { width: 100%; margin-top: .3rem; accent-color: #6d597a; }
.stats { display: flex; gap: .7rem; margin: .7rem 0; flex-wrap: wrap; }
.stat { flex: 1; min-width: 130px; background: #f6f4f8; border: 1px solid #e6e2ee; border-radius: 10px; padding: .6rem .8rem; }
.stat span { display: block; font-size: .8rem; color: #777; }
.stat b { font: 800 18px ui-monospace, monospace; color: #6d597a; word-break: break-word; }
.cmp { font-weight: 700; font-size: .95rem; min-height: 1.4em; margin-bottom: .8rem; }
.bars { display: flex; align-items: flex-end; gap: 3px; height: 110px; }
.bar { flex: 1; background: linear-gradient(180deg,#6d597a,#9a8aa8); border-radius: 3px 3px 0 0; min-height: 2px; position: relative; }
.bar small { position: absolute; bottom: -16px; left: 0; right: 0; text-align: center; font: 600 8px ui-monospace, monospace; color: #888; }
.note { font-size: .85rem; color: #999; }

// Code not found

Notice how fast it leaves reality behind: a laptop handles a handful, a supercomputer maybe ~50, and then the numbers rocket past all the storage on Earth and beyond the atoms in the cosmos — for a system you could hold in your hand. That gap is why simulating nature is hard.

Quantum hardness has a different flavor from the rest of this site:

• The state is exponential. n qubits live in a space of $2^{n}$ complex amplitudes. Just writing down the state is already infeasible for large n — before you compute anything.
• It's believed classically intractable. General quantum simulation sits in BQP (what quantum computers can do efficiently), widely believed to be beyond classical polynomial time. This isn't NP-hardness; it's a different mountain.
• Quantum computers do it naturally. A quantum computer's own state is $2^{n}$ amplitudes — so it represents and evolves quantum systems without the blow-up. That's Feynman's insight: fight quantum with quantum.
• Clever classical tricks help — sometimes. Tensor networks efficiently handle weakly-entangled systems; quantum Monte Carlo works when there's no "sign problem." But strongly-entangled systems defeat them.
• The near-term tools. Algorithms like VQE (variational quantum eigensolver) and quantum phase estimation target chemistry on today's and tomorrow's quantum hardware.

So quantum simulation is the problem that motivated quantum computers — the clearest case where a quantum machine should leave classical ones far behind.

The payoff for cracking quantum simulation would be enormous:

• Drug discovery: modeling how molecules bind, to design medicines faster.
• Materials science: superconductors, magnets and novel materials by design.
• Batteries and catalysis: better energy storage and cleaner chemical processes (e.g. fertilizer production).
• Chemistry: reaction rates and properties computed from first principles.
• Fundamental physics: understanding high-temperature superconductivity and exotic states of matter.

These are exactly the problems where classical methods hit the $2^{n}$ wall — and where useful quantum computers could matter most first.

Quantum simulation is hardness written into the fabric of physics. Nature keeps track of $2^{n}$ amplitudes effortlessly; our classical machines drown trying to copy it. It's not that we lack a clever algorithm — the information itself is too vast to store.

That's why this problem is special: it didn't just reveal a limit, it inspired a new kind of computer. Feynman's response to "nature is too hard to simulate" wasn't to give up but to build a machine made of the same quantum stuff. If quantum computers ever deliver on their promise, quantum simulation — chemistry, materials, medicine — is likely where they'll prove it first.