We are thrilled to announce that IonQ Forte, which has 32 physical qubits, is the first known quantum computer with a demonstrated #AQ of 29.
The #AQ benchmark provides a meaningful way to measure the real-world usefulness of a quantum computer, and it has been embraced by our customers and partners as a practical metric for selecting the right quantum system for each application.
IonQ has set an ambitious #AQ technology roadmap. When we introduced our IonQ Aria quantum system in February 2022, we reported an #AQ of 20. Since then, we were able to improve IonQ Aria's performance to achieve #AQ 23 followed by #AQ 25. Now, we have attained #AQ 29 with IonQ Forte–achieving our most technical goal for 2023 seven ahead of our roadmap schedule.
The achievement of #AQ 29 is a significant milestone for IonQ and one that validates the hard work and dedication of our talented team. In addition, it clearly demonstrates that IonQ’s trapped-ion quantum technology can scale to plan.
How Do We Calculate AQ?
#AQ is based on a set of familiar quantum algorithms that span a range of use cases and are representative of the workloads quantum researchers are employing in their work.
To arrive at an #AQ score and ensure our metric represents useful information for application development, we use the most relevant subset of algorithms identified by the Technical Advisory Committee on Standards and Performance Metrics at the Quantum Economic Development Consortium (QED-C), specifically for benchmarking quantum computers, and run them in increasing order of difficulty until the quality of the results falls below the target threshold. The most complex circuits in terms of the number of qubits and number of gates determine the #AQ.
We then distill the results of those algorithmic tests into a single metric - the #AQ number - which makes it easy for users to compare the relative performance of quantum computers. The #AQ result is a reflection of both how many qubits in a quantum computer can be applied to a quantum circuit, and how deep a circuit can be run on those qubits for these representative applications.
We recently observed announcements of quantum processors with hundreds of qubits, which can represent an impressive feat of engineering in those systems. However, the number of qubits is far from being a true measure of computational power for a quantum computer. The usefulness of a given quantum computer depends on a large collection of characteristics and the interactions among them. Making any claim of usefulness requires benchmarking using real-world algorithms and that's exactly what #AQ offers.
How does IonQ Forte improve #AQ?
IonQ Forte achieves the #AQ 29 milestone via several advancements:
High gate fidelity. IonQ Forte’s laser system can individually steer laser beams to ion locations, leading to more precise qubit control.
Fully software configurable. IonQ Forte enables configuration via software to dynamically address and operate ions.
More usable qubits. IonQ Forte’s architecture allows for addressing and operating more qubits than prior systems.
IonQ Forte's high-achieving application benchmarks are enabled by high-quality qubit operations. Component-level benchmarks indicate IonQ Forte offers an average 1-qubit gate error rate of 0.03% and an average 2-qubit gate error rate of 0.35%. For customers, this means the ability to run more operations with fewer errors than was previously possible.
Why these improvements matter
The improvements to IonQ Forte mean that researchers will be able to bring to bear vastly more computational power on the problems they’re solving. This represents the ability to run wider (up to 29 qubits) circuits with deeper gate counts (up to ~800 entangling gates) in these benchmark applications.
IonQ Aria and Harmony Upgrades
In addition to IonQ Forte, we have also brought improvements to our previous generation systems, IonQ Harmony and IonQ Aria.
IonQ Harmony
Released in 2019 with an #AQ 6, the quantum system has now reached #AQ 9, signifying an 8x performance increase.
Quantum noise, qubit decoherence, and quantum control inaccuracies as imagined by Midjourney AI
We also released a new error mitigation strategy, debiasing, one of the many tools we used to achieve this significant performance upgrade. Debiasing is a technique that allows us to reduce specific systematic errors in system performance with no additional qubit or gate overhead.
Combined with a post-processing technique called sharpening, algorithmic performance can be even further amplified for many common algorithms — sharpening acts as a filter to further increase the contrast between the expected results and residual noise. Both techniques are now available for all IonQ systems.
IonQ Aria
In addition, IonQ Aria is available to customers and partners worldwide, and the system has already seen significant commercial use. Hyundai Motor Company has been using IonQ Aria to perform image classification for autonomous driving and quantum calculations to study battery chemistry. And GE Research has used IonQ Aria to successfully demonstrate the high potential of quantum computing for risk aggregation.
Ready to get started with IonQ’s quantum computers? You can visit our registration page to get started today with the right system for your needs.