When Quantum Possibilities Collide

How interference helps quantum computers find useful answers

Two mornings a week, a group of us paddle our 18-foot surf skis south from Cottesloe Beach towards the breaks at Dutchies or Cables. When the swell is up, we punch southwest towards the sand and reef breaks.

When the surfing is done, we ride the swells home. We call these runs skates and the skis gather speed as they race down the face of each swell.

As we approach the rock groyne (breakwater for US readers) at Cottesloe main beach, the ride becomes unsettled. Earlier swells have struck the rocks, reflected and returned to slap against our skis.

There is a pattern in the chaos. It is the same phenomenon a quantum computer uses to turn many possibilities into a useful answer.

When the crests of two waves meet, the water rises and when a crest meets a trough, they flatten each other out. Physicists call this interference and it is the third concept in this Making Sense of Quantum series after superposition and entanglement.

Creating the possibilities

Quantum computers work with massive possibilities that are enabled by superposition, which allows a qubit to hold multiple states, and entanglement that links qubits so they behave as parts of one system.

When the qubits are read, or measured as the physicists say, a result is delivered.

If the measurement is done at random the result will be random. So what makes a useful outcome more likely to appear?

The waves do the work

The answer is interference.

Each possible state of a qubit system has an amplitude, a quantity with a size and a phase. Amplitudes form a pattern similar to a wave, much like the height and timing of the swell and backwash at the groyne.

Some algorithms test for the conditions that a useful answer must satisfy and adjust the amplitudes so that those associated with correct answers reinforce each other and unhelpful possibilities cancel out. This is interference.

By the time the qubits are measured, unhelpful possibilities have been suppressed and a useful answer is more likely to appear.

Knowing when to stop

As with a lot of things in life, the key is knowing when to stop.

One example is Grover’s algorithm, which runs a series of operations designed to maximise the probability of a useful result. Run too few iterations and the probability remains low. Run too many and it begins to fall again. The algorithm therefore runs for a calculated number of iterations before the qubits are measured.

As the system is dealing with probabilities, the process may be repeated multiple times to build confidence in the result.

Why the possibilities are not enough

It is often said that quantum computers process all possible answers simultaneously. The real power lies in using interference to arrange the possibilities so that a useful answer is more likely to appear when the qubits are measured.

This is why a quantum computer is not just faster than a classical computer. Quantum works in a fundamentally different way that, for specific problems, allows it to reach answers in far fewer steps than classical methods.

Where quantum makes a difference

While superposition and entanglement provide a range of possible results, it is interference that converts those possibilities into a useful outcome.

For many everyday tasks, this process is unnecessary and using a classical computer is a better option.

For boards and executives exploring quantum solutions, the key question is whether an algorithm exists or could be developed to solve an important organisational problem that is beyond the reach of classical systems.

Potential candidates include analysing financial risk across vast numbers of scenarios, modelling molecules to develop new drugs and optimising complex transport, energy and supply networks.

Here in Perth, researchers tackling these issues include those at the University of Western Australia’s Centre for Quantum Information, Simulation and Algorithms. The Centre and its counterparts around the world are working with industry to turn quantum principles into practical applications.

At Cottesloe, waves meet at the groyne. Sometimes the waves form up into bigger waves, at other times they flatten out in a dance determined by the rocks and the swell. Quantum algorithms use the same principle, but deliberately shape the pattern so that a useful answer is the one most likely to stand up when you look.

Next week: Why Are Quantum Computers So Difficult to Build?

Creating quantum possibilities is difficult. Keeping them alive long enough to be useful is harder.

Notes from the Intersection is a free publication exploring how emerging technologies, ideas and culture are reshaping decisions, organisations and society