Coming Soon: Quantum Computing on Your Desktop PC?

Today has been pretty dull in the world of political news, so let’s continue trawling other parts of the global knowledge ecosystem for interesting tidbits. This one looks potentially important:

For decades, researchers have been trying to build a computer that harnesses the enormous potential of quantum mechanics. Now engineers from the University of New South Wales (UNSW) in Australia have overcome the final hurdle, by creating a quantum logic gate in silicon — the same material that today’s computer chips are made from.

The newly developed device allows two quantum bits — or qubits — to communicate and perform calculations together, which is a crucial requirement for quantum computers. Even better, the researchers have also worked out how to scale the technology up to millions of qubits, which means they now have the ability to build the world’s first quantum processor chip and, eventually, the first silicon-based quantum computer.

Quantum computing is sort of like fusion power: constantly right around the corner, but never quite there. The fundamental problem is that qubits suffer from decoherence unless they’re kept completely isolated from their surrounding environment, which is pretty tough since they also need to communicate with other qubits in order to be useful. Researchers have gotten a lot better at controlling qubits in recent years, but as the UNSW paper points out, this has required the use of fairly exotic materials: “single photons, trapped ions, superconducting circuits, single defects or atoms in diamond or silicon, and semiconductor quantum dots.”

By contrast, a two-qubit logic gate that can be implemented in silicon using standard lithographic techniques is a whole different matter. If this turns out to be for real, chips containing thousands or millions of qubits are finally within practical reach.

This would be very cool, though only for a certain subset of problems amenable to massive parallel processing. This is inherent in the difference between standard computers and quantum computers. A standard computer with, say, 50 bits, can be in any one of 250 states at a single time. That’s about a quadrillion states. This state changes with every beat of the computer’s internal clock, and eventually you get an answer to whatever problem you’ve programmed the computer to solve. By contrast, a quantum computer with 250 qubits can be in a quadrillion states simultaneously thanks to an aspect of quantum weirdness called superposition. Once you set up the program, you just collapse the quantum state and the answer is spit out instantly.

This is not the kind of thing you’d use to write an iPhone app. But it could be used to break some public-key encryption systems. It might also be useful for things like modeling protein folding, which is fundamentally a quantum problem that requires a tremendous amount of computing time using traditional computers. There’s also potential for exponentially faster database queries.

And one other thing: it’s possible that large-scale quantum computing could lead to breakthroughs in emulating human thought processes and speeding up the creation of artificial intelligence. Maybe.

Anyway, it’s fascinating stuff, and it seems as if useful quantum computing may be finally getting within reach. If it does, it would blow away Moore’s law for certain kinds of problems—possibly many more than we think once we get the hang of writing a whole different kind of code. In a few years, maybe we’ll even get customer support voice recognition systems to work properly.