Around six years ago I started publicly warning everyone that quantum computers capable of cracking encryption were about ten years away. All the initial technical hurdles were vanishing, and I extrapolated a spooky physics version of a new Moore’s Law manifesting.
I haven’t backed down on my initial hot take. Not by one qubit.
I’m publicly still on track, warning of a now three to five-year window in which everything must change. 2017 saw the birth of multiple ~50-qubit machines. Google is now pushing into the 70s. These may sound like low numbers for a crypto apocalypse, but if you understand how simulating quantum processors on classical hardware works, we’re past the event horizon here. You may hear the phrase “quantum supremacy”. For now, this means: when a quantum computer exists that classical computers can’t simulate.
2018, therefore, reigns supreme. We finally have machines that cannot be flukes or statistical noise. These are real machines proving the wondrous power of using zero and one at the same time, instead of having to make a limiting binary choice.
I’m not bragging, though. I get things wrong, and I’m delighted about one thing I didn’t see happening. For some reason, I thought the initial quantum systems would stay buried in labs, with very tedious interfaces analogous to the days of punch cards. As a Linux geek, I should have instead been thinking of the unlimited power of open source. Rather than hide these machines away, companies are sharing their use over the cloud, and providing open-source development kits to code on them!
All the world’s 500 top supercomputers run Linux, so I shouldn’t have been surprised that the first coding interfaces to the next generation of super systems would also be open source. IBM allows users to log in and run simple programs on up to 16-qubit processors for free with its Q Experience (20-qubit systems available for customers, with the 50-qubit monster not accessible online). Rigetti is following suit, with Microsoft looking to move from the realm of simulators, too. No one’s cracking encryption yet, but as I’ll point out in my talk at RSAC, real scientific problem solving and paper writing is happening.
The first time the computing revolution was upon us, people had to practically break into locked rooms to experiment with the behemoths. Now you can run real quantum programs either on simulators on your computer or processors with physical qubits via the cloud. The first time I created a qubit over such an interface floored me. Experience it!
It’s time to get in on the ground state here — to be ready for the next generation of computing that is not only coming, but necessary. We’re reaching the limits of miniaturisation in classical computing. A new architecture is a must.
When our bank of quantum algorithms goes from dozens to hundreds, we’ll be able to do amazing things with even less power and time. Make sure you’re a part of tomorrow's supercomputing. The open source invitation is real. Today.