Jeff Bezos and CIA Making Bet on Quantum ComputerWhen you are Jeff Bezos, you have so much money you have to look for projects to bet on. Back in the 1990's, Bezos founded a small web-based bookseller called Amazon. It had a bit of success. Since then, he's put money on tablets, e-readers, rocket companies, even raised the first stage of Apollo 11's Saturn V launcher. His next bet, however, might just prove to be the one people remember him the most for.
Photo of a chip constructed by D-Wave Systems Inc.
Bezos is betting on quantum computers, along with the CIA's investment arm In-Q-Tel. They, along with a few smaller investors, are betting $30 million that a small company out of Vancouver called D-Wave.
D-Wave has made waves in the past, being the first to offer commercial quantum computers. But the caveat is that their last generation of devices were only pseudo-quantum (a distinction I'll never understand) and couldn't even match traditional computers. Also, that last wave of machines tarnished their reputations, something they hope to rectify.
D-Wave has been on a bit of a roll, lately. Last May the company published a paper in Nature that critics felt proved the D-Wave quantum chips did, in fact, have the properties needed to back up the company's claims. Google has been using their hardware extensively, with artificial intelligence researcher logging onto the machines just to see what they are like. Lockheed Martin paid $10 million to develop the technology into a tool to find an annoying bug in the design of the F-35 fighter.
So, how does it work?Keep in mind, here, that I am not a quantum physicist. These are thorny matters, and we're on the edge of science and technology, here. But my understanding goes something like this:
Start by pumping super-cooled fluids through the contraption, in a desperate plea to get the device down to temperatures barely imaginable. Shield the insides from magnetic fields, as any could completely destroy the effect. Rather than build logic out of silicon, do it from niobium. The loops of niobium trap electrons, which can spin either one way or the other, representing an on, off, or both state. D-Wave currently used 512 of them, so the chip has 512 quantum bits made like this, which is quite small when compared to traditional computing. Then lay out those qubits in such a way that it can compute specific types of problems.
Quantum computers rely on the quantum mechanical nature of small bits to do computation. Most traditional computer elements can be on or off, but quantum computers can be both, massively in parallel. The qubits enter a state called 'entanglement,' in which they functionally explore every possible way of calculating the problem, before settling down on the correct answer. The interesting thing is that computational time doesn't scale like classical computers do. We're not even talking about magnitudes of difference, here, but magnitudes of magnitudes of differences. Problems that would normally take longer than the entire life of the universe to calculate on our existing machines suddenly can be done in minutes, if the quantum chip is big enough, and problems that there are no natural ways to solve suddenly become feasible.
That is why quantum computers are the holy grail of computing. All the traditional bottlenecks disappear in a way that hasn't happened since the first computer was made.
But first D-Wave needs to prove that its chips are quantum.