Original version With this story appeared in Quanta Magazine.
It’s not effortless to study quantum systems – collections of particles that follow the counterintuitive principles of quantum mechanics. Heisenberg’s uncertainty principlecornerstone of quantum theory, says that it is impossible to simultaneously measure the exact position of a particle and its speed – this is quite crucial information for understanding what is happening.
To study, say, a specific set of electrons, researchers have to be clever. They can take a box of electrons, poke it in different ways, and then take a snapshot of what it looks like at the end. In this way, they hope to reconstruct the internal quantum dynamics at work.
However, there is a catch: you cannot measure all system properties at the same time. So they repeat. They will start with their system, poke, and then measure. Then they will do it again. In each iteration they will measure some fresh set of properties. Build enough snapshots and machine learning Algorithms can support reconstruct the full properties of the original system—or at least get really close to them.
It’s a tedious process. But theoretically quantum computers could support. These machines, which operate according to quantum laws, have the potential to model the operation of quantum systems much better than ordinary computers. They can also store information not in classical binary memory, but in a more convoluted form called quantum memory. This allows for much richer and more precise descriptions of particles. This also means that a computer can store multiple copies of a quantum state in its working memory.
A few years ago, a team from the California Institute of Technology demonstrated that some algorithms that apply quantum memory require exponentially fewer snapshots than algorithms that do not apply it. Their method was a huge advance, but required a relatively vast amount of quantum memory.
This is something of a deal breaker because, from a practical standpoint, quantum memory is difficult to come by. A quantum computer consists of quantum bits connected together called qubits, and qubits can be used for computation or memory, but not both.
Now two independent teams have come up with ways to deal with much smaller quantum memory. In the first one paper, Sitan ChenHarvard University computer scientist and his co-authors have shown that just two copies of a quantum state can exponentially reduce the number needed to take a snapshot of a quantum system. In other words, quantum memory is almost always worth the investment.