Google’s quantum computer is nearly 1.5 trillion times faster than existing supercomputers, according to a new paper published on 23 October in Nature (1). The paper — officially released just five weeks after a leak of draft outlining the achievement was posted on the NASA website — claims their Sycamore processor can perform a task in 200 seconds that would take a supercomputer approximately 10,000 years to complete — an experimental realization of quantum supremacy (1).
Quantum supremacy is considered proof that quantum computers can outperform conventional computers. And something experimental physicist Dr John Martinis has been working towards for around two decades.
Google’s quantum computer, developed and built by Martinis and his team at Google’s facilities in Mountainview, California and the University of California Santa Barbara, runs on quantum chip comprised of 54 entangled quantum bits (qubits). Each qubit has its own superconducting circuit.
In their proof-of-concept experiment, the researchers showed that the computer is capable of performing a fixed set of operations that entangles 53 qubits into a complex superposition state (one of the 54 qubits was broken). This was achieved by applying a sequence of tens of control pulses to each qubit in a matter of microseconds.
The quantum advantage
Whereas digital computers use data called bits that can only exist in one of two binary states (1 or 0), qubits can exist in multiple states at once — something called a superposition. And these so-called quantum mechanical states are what gives qubits their unique — and potentially powerful properties — allowing them to do much more than conventional bits.
Of course, exploiting these wave-like quantum states is much more complicated. Qubits must be grouped into collections of superpositions to perform complex tasks. But qubits are incredibly sensitive it becomes much harder to maintain their fragile states when they are linked, which is why quantum computing has remained so elusive. Moreover, millions of qubits could be needed for an everyday machine.
A random number generator
The algorithm was chosen as a way of demonstrating the strengths of their quantum machine. In other words, to test its ability to hold and rapidly manipulate a huge amount of complex, unstructured data. But the researchers have essentially created a lighting-fast random number generator, with limited yet potentially useful applications in the future. For example, quickly generating secure and totally random encryption keys.
The next steps will be using their quantum computer to perform a task that cannot be solved in any other way and to demonstrate quantum supremacy in an algorithm that uses error correction, which corrects for noise-induced errors that could otherwise botch a calculation. The latter will be essential if quantum computers have any hope of being used at scale.
So it’s also important to remember that while this latest milestone is definitely an incredible feat, everyday quantum computers are still decades away. But other big-tech firms like Intel, Microsoft, and IBM, as well as many many startups, are also in the race towards a viable quantum computer, perhaps, a testament to its undeniable potential.
(1) Arute, F. et al. Quantum supremacy using a programmable superconducting processor. Nature (2019). DOI: 10.1038/s41586-019-1666-5