During the last few years, IBM has devoted a special interest to the quantum computingwith a ambitious roadmap that we have known for a couple of years.
The company recently presented advances in quantum error mitigation, in order to approach quantum hardware that is increasingly fault-tolerant and therefore more efficient.
Error Mitigation for Efficient Quantum Computing
Although we have recently seen many advances in quantum computing, its practical implementation still has many pending tasks. There is a consensus that a large fault-tolerant quantum processor needs to be built before any of the proven superpolynomial acceleration quantum algorithms that have emerged in various research projects can be implemented.
According to IBM in its reportrecent advances in quantum error mitigation techniques allow us to trace a smoother path towards this goal. “Along this path, advances in qubit coherence, gate fidelity, and speed immediately translate into a measurable advantage in computing, similar to the steady progress historically seen with classical computers.”says the company regarding its latest developments.
Today’s quantum hardware, IBM explains, is subject to different sources of noise, the best known being qubit decoherence, individual gate errors, and measurement errors. These errors limit the depth of quantum circuitry that we can implement. However, even for shallow circuits, noise can lead to erroneous estimates.
Thanks to the application of quantum error mitigation techniques, these teams can count on a collection of tools and methods that allow evaluating accurate expected values of shallow and noisy quantum circuits, even before the introduction of fault tolerance.
Recently, IBM introduced and implemented two error mitigation methods general purpose, called “zero noise extrapolation” (ZNE) and “probabilistic error cancellation” (PEC).The ZNE method cancels subsequent noise orders that affect the expected value of a noisy quantum circuit by extrapolating the measurement results to different noise levels. More recently, theoretical and experimental advances have shown that PEC can already enable noise-free estimators of quantum circuits on noisy quantum computers.
IBM stresses that the path to quantum advantage will be driven by improvements in the quality and speed of quantum systems, as they scale to address increasingly complex circuits. Among its contributions to this goal, the recent introduction of its 127-qubit Eagle processor stands out, which takes the scale of these processors beyond the limits of exact classical simulation.
They also highlight the application of a metric to quantify the speed of quantum systems (CLOP), with which they demonstrated a 120-fold reduction in the execution time of a molecular simulation. Another performance milestone is marked by the coherence times of its superconducting qubits, which exceeded 1 ms.
The company reports that these improvements have been extended to its largest quantum processors. Graphing their input, they highlight that the 65-qubit Hummingbirds have seen a 2- to 3-fold improvement in coherence, allowing for higher-fidelity gates. They also report that in their latest processor, the Falcon r10, the gate errors of two qubits have been reduced to below 0.1%.
The division of the company will continue to focus on enhancing its advances. “At IBM Quantum, we plan to continue developing our hardware and software with this path in mind. As we improve the scale, quality, and speed of our systems, we expect to see (…) quantum runtime improvements for circuits of interest.”they noted in their ad.
