Google is laying its claim to quantum supremacy. In a paper published in the scientific journal Nature, the search giant announced that its Google AI Quantum team developed a processor capable of completing computations that were otherwise unfathomable for existing supercomputers to solve. The achievement means Google has created the most powerful quantum computer in the world and have achieved quantum supremacy — through the accomplishment is already being challenged by one of its biggest competitors.
To understand the achievement Google is laying claim to, it's important to know that quantum computing has largely been theoretical up to this point. Most computers, from your phone to your laptop to even powerful supercomputers that can complete trillions of computations per second, rely on traditional bits to store and manipulate information. Bits store information in binary, as 0s and 1s. Quantum computers do not have this restriction, as they use a mechanical phenomenon known as quantum bits or qubits to store information. Qubits can hold values of 0 and 1 or both simultaneously.
This is made possible because quantum computing relies on microscopic particles like electrons and photons, which can take on different states simultaneously. This isn't possible typically, because when an electron or photon is observed, we then know its location and polarity. But prior to it being observed, it is possible for it to occupy multiple states at once, known as superposition. It's the very principle that the Schrödinger's cat thought experiment attempts to explain: until you look at something, you never truly know its state.
Working with those principles, Google built its Sycamore processor, which is capable of creating quantum states on 53 qubits. In order to test the capabilities of the Sycamore processor, Google tested the computing power versus a traditional computer. Here's how it works: the quantum computer is tasked with executing a random collection of computations, called a quantum circuit. That same process is simulated on a standard computer. When the quantum computer is still capable of completing the circuits but the classic computer can no longer keep up, the quantum computer has achieved quantum supremacy. For Google's Sycamore processor, it was capable of sampling a quantum circuit one million times in about 200 seconds — a task that would take a classical supercomputer about 10,000 years to complete, according to Google's simulations.
At least, those are Google's claims. The company has already been challenged by IBM, which just so happens to be one of the biggest competitors in the realm of quantum computing. In a blog post published shortly after Google's claims to quantum supremacy were made public, IBM argued that, if given the ideal conditions including the necessary amount of disk storage, a supercomputer like IBM's own Summit could complete the same computations as the Sycamore processor in two and a half days rather than 10,000 years. That's still a significant difference than the about three and a half minutes that it takes Google's processor to do the task, but it's a far cry from quantum supremacy — a title that can only be claimed when it is no longer possible for a traditional computer to feasibly complete the task.
Whether Google actually achieved quantum supremacy or not is certainly important for those most invested in these developments, but there's a clear win for Google in this debate between computing giants. What is important here is that the company's quantum computer works, and it works exponentially faster than even the world's most powerful traditional supercomputer. It's one step toward making quantum computing a reality, which would open all kinds of new doors.
It could accelerate the development and advancement of artificial intelligence, enable quantum chemistry that can precisely model molecular interactions, and vastly improve modeling techniques for everything from weather forecasts to finances. For now though, quantum computing is still largely theoretical, even if Google did successfully achieve quantum supremacy as they claim. The calculations that the Sycamore processor had to complete to reach this goal have essentially no practical use other than to test the computational power of the processor. Consider this the test page that your printer spits out after you first plug it in — the ink comes out, now what are you going to do with it? We're still a long way from quantum computing being used for practical purposes, but the achievement moves us one step closer to that being a reality.