Building a slide deck, pitch, or presentation? Here are the big takeaways:
- Google has announced the release of the 72-qubit square array Bristlecone quantum processor, which the company believes is adequate to demonstrate quantum supremacy.
- Bristlecone is the evolution of Google's prior 9-qubit linear quantum processor, which had error rates of 1% for readout, 0.1% for single-qubit gates, and 0.6% for two-qubit gates.
Google unveiled its Bristlecone quantum processor Monday at the annual American Physical Society meeting in Los Angeles. The traditional gate-based processor is intended as a test platform for researching error rates and scalability of Google's approach to implementing qubits in quantum systems.
One of the primary encumbrances to building a quantum computer is the comparatively high error rate that quantum processors are prone to. These systems require extremely precise environmental controls, as fluctuations in the operating environment can cause errors in a given computation to occur. Presently, reliably repeating operations in computers with a relatively small number of qubits is the first challenge to overcome before quantum computers can be scaled beyond tens of qubits.
Google's previous design was a linear nine-qubit system, which the company was able to refine to relatively low error rates. According to Google, it achieved best case error rates of 1% for readout, 0.1% for single-qubit gates, and 0.6% for two-qubit gates. The Bristlecone processor scales the same design for coupling, control, and readout in a square array of 72 qubits.
SEE: IT leader's guide to the future of quantum computing (Tech Pro Research)
On Bristlecone, Google intends to facilitate the development of quantum algorithms on real hardware rather than simulated hardware, as well as investigate first and second level error correction schemes, and attempt to demonstrate quantum supremacy in the future.
Quantum supremacy is the threshold at which a quantum computer performs in a way that traditional computers are practically incapable of doing (or, otherwise, the use of a traditional computer for a given task would be infeasible.) As a basic type of benchmark, quantum supremacy can be established by executing Shor's algorithm to factor integers with a superpolynomial speedup compared to the fastest algorithm executable on a traditional computer.
Google plans to use an apparently original test to establish quantum supremacy. According the announcement:
Our theory team has developed a benchmarking tool for [quantum supremacy]. We can assign a single system error by applying random quantum circuits to the device and checking the sampled output distribution against a classical simulation. If a quantum processor can be operated with low enough error, it would be able to outperform a classical supercomputer on a well-defined computer science problem.
Google believes that quantum supremacy is possible at 49 qubits, with a circuit depth higher than 40, and a two-qubit error rate less than 0.5%. That said, Bristlecone is a transitory, proof-of-concept processor, which, according to the researchers, "requires harmony between a full stack of technology ranging from software and control electronics to the processor itself. Getting this right requires careful systems engineering over several iterations."
Of note, the achievement of quantum supremacy would be one step closer to breaking public-key cryptography, as schemes such as RSA are built around the notion that factoring large numbers is practically too difficult. That said, this type of factorization requires a quantum computer with thousands of qubits, which is beyond what is presently available.
Other organizations are also researching quantum computing techniques. IBM has a 50-qubit machine in development, as well as a cloud service allowing researchers to test quantum algorithms. D-Wave advertises their line of quantum computers as having thousands of qubits, though these systems are designed specifically for quadratic unconstrained binary optimization. D-Wave's definition and measurement of qubit does not strictly conform to the standard used by academics and other companies, making the comparison inexact.