As we approach the physical limits of shrinking the components on computer chips, it’s getting harder to attain more computing power. But quantum computers—which rely on the strange properties of atom-sized logic gates—could have exponentially more power than even today’s most advanced supercomputers. These machines could be used to develop new drugs, crack codes, or potentially even create true artificial intelligence.
IBM has been working on quantum computers for decades, and now it has the support of the US intelligence agencies. On Nov. 8, IBM announced that IARPA, the intelligence community’s research arm, had awarded the company a multi-year grant to continue its research into building quantum computers.
Classical computers—everything from the giant mainframes that used to take up entire rooms, to the smartphone in your pocket—consist of billions of tiny transistors. Those are switches, essentially, that can be turned on or off, represented on computers as 1 or 0. Every, tweet, text, document, and photo is composed of a series of those 1s and 0s, known as bits. In quantum computing, a transistor consists of a single atom, which can exist as a sort of combination, or “superposition,” of 1 and 0 at the same time. These quantum bits, or qubits, in essence allow a computer to perform multiple parts of a calculation at once, making it far more powerful.
But that’s assuming you can get enough qubits working in one place. These qubits are much less stable than traditional bits, and need to be isolated in large machines that can cool them to almost absolute zero. IBM has a few such machines, and the most powerful contains a mere eight qubits.
Right now, IBM is working on correcting errors that are thrown up during its quantum computing tests—due to the fragility of the qubits—and assuming it’s able to wrangle thousands of qubits in one place, all computing as desired, it would potentially have a computer far more powerful than one with a similar number of traditional bits.
One key area where quantum computers could have an impact is encryption and security. Governments could use them to crack even the strongest current forms of digital encryption, or create new forms that are unbreakable. This is likely one reason why IARPA is interested in quantum computing. The agency has been working with other companies, like D-Wave, to develop these machines. (Though some in the field aren’t entirely sure if D-Wave’s computers are really quantum computers, researchers at Google—which bought a D-Wave machine in 2013—have just posted some new findings that they say show “a 100-million-fold speed-up” on a certain, very specific kind of problem.)
IBM outlined its quantum-computing roadmap yesterday (above) but told Quartz, when we visited the lab earlier this year, that it could well be many years, if not decades, before a viable quantum computer is built. But a patient, methodical, and practical approach to research is what has helped the company develop world-changing technologies in the past.