Quantum Computing News

Analog Counterdiabatic Quantum Computing (ACQC)

In a significant step toward achieving practical quantum advantage, a team of researchers from Kipu Quantum, QuEra Computing, and Pasqal has demonstrated a novel method to drastically accelerate the performance of neutral atom quantum processors. The article presents Analog Counterdiabatic Quantum Computing (ACQC), a method that reduces the inherent faults of standard quantum evolution to produce high-quality solutions for challenging industrial issues.

Using up to 100 qubits, the researchers were able to obtain a threefold speedup in convergence time and solution quality by applying this strategy to the Maximum Independent Set (MIS) problem, a well-known combinatorial issue. For “noisy” near-term quantum devices, which are frequently constrained by non-adiabatic errors and short coherence durations, this result represents a turning point.

The Race Against Time: Overcoming Coherence Limits

Optimizing logistical routes, scheduling tasks, and creating effective telecommunications networks are just a few of the computationally challenging challenges that quantum computers are intended to handle. Adiabatic evolution, in which a system starts in a known ground state and gradually changes into a final state that encodes the solution to a problem, is a cutting-edge method in analog quantum computing.

The coherence window is a significant obstacle to this process, though. If evolution is too quick, “non-adiabatic errors” occur when the system jumps out of its optimal route, lowering the fidelity of the final result; if development is too slow, noise destroys the quantum state.

The researchers observed that “these errors limit scalability in state-of-the-art adiabatic protocols.” To get around this, the group used counterdiabatic (CD) protocols, which block undesired transitions between energy states by adding an auxiliary term to the system’s Hamiltonian. Although CD terms have been predicted for a long time, they were previously hard to compute for big systems or required intricate “many-body” interactions that are difficult to accomplish with present hardware.

A Tailored Solution for Neutral Atoms

The unique ACQC protocol created especially for neutral atom processors is an innovation. To encode information, these processors employ arrays of atoms held in optical tweezers, taking advantage of the Rydberg blockade phenomenon, which occurs when an excited atom stops its neighbors from becoming excited.

The researchers created an analytical technique to compute these counterdiabatic corrections without requiring knowledge of the entire energy spectrum or resource-intensive variational iterations. The ACQC technique directs the system along a “shortcut” to the solution by dynamically modifying the driving lasers’ Rabi frequency, detuning, and laser phase.

The ACQC method employs totally analog, continuous control, in contrast to digital quantum processors that use discrete, error-prone gate sequences to implement CD protocols. This is perfect for the limited coherence of today’s hardware since it enables higher-quality solutions in shorter amounts of time.

Benchmarking Performance: 100 Qubits and Beyond

Real-world hardware, such as Pasqal’s Orion Alpha CPU and QuEra‘s Aquila device, was used to test the efficacy of ACQC. Finding the greatest set of non-adjacent nodes in a graph is the MIS problem, which the team concentrated on. This challenge has obvious applications in computational biology, resource allocation, and network resilience.

Among the experiments’ main conclusions are:

• Rapid Solutions: With an evolution time of only 1 microsecond, ACQC clearly outperformed conventional adiabatic techniques in the “fast quench” region.
• Higher Success Probabilities: ACQC had a 60% success rate on a 15-node simulation, while a typical linear adiabatic procedure only had a 27.8% success rate.
• Scalability: Even at very short durations, the team was able to maintain a 6–8% advantage in solution quality over smooth adiabatic schedules by successfully handling 100-node graphs on the Aquila processor.
• Significant Gains: Compared to conventional approaches, ACQC produced a 5x improvement in the approximation ratio during short evolution times on a 27-node graph employing Pasqal hardware.

The capacity to find high-quality solutions fast is crucial, the researchers noted, even though performance converges for longer evolution times. “The fast-quench feature can be used in sequential processes where one part of the computation time is used to prepare a state and the other to perform high-fidelity computation within the coherence time,” the report stated.

Industry Relevance and Future Outlook

There are significant ramifications for the industry. The MIS issue this study tackles relates to intricate logistical problems, such as choosing non-conflicting subsets of financial assets or determining the most reliable communication networks. ACQC delivers a “plug-and-play” improvement for existing and upcoming devices by offering a broad analytical method that doesn’t require extra processor optimization.

The team anticipates that future hardware developments like single-qubit addressability and local detuning will help ACQC. These characteristics would make it possible to develop even more sophisticated CD techniques, which could lead to previously unheard-of levels of precision in the study of critical dynamics in spin systems.

It will be essential to use techniques like ACQC that lessen the experimental requirements for proving quantum advantage as quantum technology continues to scale toward hundreds and thousands of qubits. In contrast to a far-off, error-corrected future, this work lays forth a potential approach for using analog quantum processors to address current industrial issues.

International specialists from the University of the Basque Country, Pasqal, and Gradiom Sárl contributed to the study, which was coordinated by Qi Zhang, Narendra N. Hegade, and Eric Michon of Kipu Quantum. The associated authors can provide the custom codes and data that support these conclusions upon request.