In This Story
Amazon (AMZN-0.66%) Web Services has unveiled its first-ever quantum computing chip that it said is a step toward building practical quantum computers at scale.
The tech giant said it designed Ocelot — a small-scale prototype chip — to test its quantum error correction architecture.
Unlike classical computers which use binary bits, quantum computers use quantum bits — or qubits — which are usually electrons, photons, or another subatomic particle. Connected together, qubits have far more processing power than binary 0s and 1s because they can be 0 and 1 at the same time in a state called superposition. The more qubits that are used, however, the more errors usually occur in a computation, which is costly to fix and makes quantum less appealing to commercialization efforts.
Amazon said it designed the chip to implement quantum error correction in an efficient and scalable way that can reduce the additional qubits needed for the process by up to 90%.
Ocelot uses cat qubits, Amazon said — a type of qubit technology that has “inherent protection against bit-flip errors,” which is one of the two types of errors quantum systems can face. With cat qubits, the company said it has shown that this kind of technology can fit on a microchip with other components of quantum error correction.
“We believe that Ocelot’s architecture, with its hardware-efficient approach to error correction, positions us well to tackle the next phase of quantum computing: learning how to scale,” Fernando Brandão, head of quantum algorithms at AWS, and Oskar Painter, head of quantum hardware at AWS, said in a statement. “Using a hardware-efficient approach will allow us to more quickly and cost effectively achieve an error-corrected quantum computer that benefits society.”
Last week, Microsoft (MSFT-0.19%) announced its first quantum computing chip, Majorana 1, built with “topological superconductors” or “topoconductors” — a novel material that enabled the company to create a new state of matter called “topological superconductivity.”
The new state of matter, which Microsoft has worked on for almost two decades, makes it possible for quantum systems to scale to one million qubits on a single, hand-sized chip, the company said, adding that it has demonstrated the world’s first topological qubit and already placed eight on Majorana 1. At one million qubits, a quantum computer could perform far more difficult calculations than traditional computers with higher accuracy.
Google (GOOGL-1.48%) also announced a quantum computing chip called Willow in December, which it said demonstrated two “major achievements” in quantum computing. Willow was able to “exponentially” reduce the rate of errors when adding more qubits and performed a computation in under five minutes. According to the company, this same computation would take Frontier, currently considered the world’s fastest supercomputer and operated by the Oak Ridge National Laboratory and U.S. Department of Energy, 10 septillion years to solve.