Quantum Elements’ Orbit boosts circuit reliability on IBM hardware

IBM Quantum users on Premium, Flex, and On-Prem plans now have access to Orbit, a new Qiskit Function from Quantum Elements that integrates dynamical decoupling, optimized transpilation, and measurement error mitigation into a single workflow, the company says. This unified approach aims to boost the reliability of quantum circuits by actively suppressing errors during execution, rather than relying on post-processing techniques. Orbit achieved a 27.51x improvement in fidelity on dynamic circuits, increasing fidelity from approximately 10% in raw circuits to approximately 0% with Orbit active.

It suggests a path toward logical-level dynamical decoupling, but does not focus on shielding logical qubits. In experiments, Bernstein-Vazirani algorithms successfully evaluated circuits reaching 70 qubits when utilizing the full Orbit workflow, compared to 26 qubits with raw circuits and 60 qubits with transpilation alone. Orbit, combined with error detection, yielded a 1.63x improvement in average entangled Bell-state logical fidelity over 55 microseconds. Long-range Bell-state fidelity also benefited with Orbit optimizing performance. The function allows for both automated optimization and granular user control, enabling experimentation and fine-tuning for specific circuits and hardware backends.

Orbit Integrates Error Mitigation with Qiskit Functions

Qiskit Functions, including Orbit, are currently an experimental feature accessible only to IBM Quantum Premium Plan, Flex Plan, and On-Prem users via the IBM Quantum Platform API, establishing a tiered access structure for this new capability. This limited initial availability suggests a strategy of focused deployment and refinement within a select user base before broader release. The function streamlines workflows by accepting a logical circuit and mapping it directly to the chosen backend during execution, simplifying the process for users accustomed to more fragmented approaches.

A practical implementation of Orbit begins with defining a quantum circuit and selecting the function through the Qiskit Functions Catalog, then establishing a baseline with a raw-circuit mode that bypasses Orbit’s optimizations. Users can then choose from a range of optimization strategies, including automated or manually selected dynamical decoupling (DD) sequences, measurement error mitigation (MEM), and optimized transpilation, allowing for granular control and experimentation.

This flexibility extends to enabling or disabling individual methods, or submitting multiple configurations to assess their combined effects, particularly in circuits with idle periods or scheduling gaps where qubits are not actively engaged in gate operations. Orbit’s impact is quantified through several metrics, including logical error probability, VQE absolute error, and logical fidelity, each offering a different perspective on circuit reliability.

Examining the results involves comparing counts retrieved from jobs run in both raw and Orbit-enhanced modes, with a specific example showing a comparison of “raw” and “orbit” labels using data from a backend named. A job with ID c0f1b99c-8dfa-4ebb-9560-d9a37c990acc was used to gather the data, demonstrating a traceable workflow from execution to analysis. The data, represented as an OrderedDict, includes detailing the raw and Orbit results for a specific hidden bitstring, allowing for a visual comparison of circuit performance.

Automated Dynamical Decoupling Enhances Circuit Reliability

This combination addresses limitations inherent in current quantum hardware by actively suppressing noise and errors during computation, rather than relying solely on post-processing correction. Selecting an effective dynamical decoupling strategy is often complex, as performance varies depending on the circuit and noise environment; Orbit automates this selection process, simplifying implementation for users. The system’s impact extends to maintaining logical qubit fidelity over time, as demonstrated in experiments entangling two logical qubits to create a logical Bell state under quantum error detection, according to IBM.

Without Orbit-style protection, Bell-state fidelity decayed to approximately 44% over tens of microseconds, but with Orbit active, fidelity remained near the upper limit, reaching approximately 95.3% across the same duration. This preservation of coherence is particularly notable given the challenges of maintaining qubit states in the presence of environmental noise. Quantifying Orbit’s effectiveness involves several metrics, including improvements in quantum Fourier transform fidelity and reductions in energy error.

For dynamic circuits, which incorporate mid-circuit measurements and classical feedforward, Orbit suppresses decoherence, leading to measurable gains in fidelity compared to raw, unenhanced circuits. The improvement is calculated as a factor comparing the energy error with and without Orbit, with values greater than one indicating a reduction in error.

Similarly, Orbit enhances QAOA success probability, increasing the likelihood of measuring a bitstring corresponding to an optimal solution on a butterfly graph. Jobs are assigned either to a raw mode, skipping Orbit’s optimizations, or to Orbit defaults, which include transpilation, dynamical decoupling insertion, and measurement error mitigation.

Orbit Achieves 27.51x Fidelity Gains on Dynamic QFT Circuits

Quantum circuits executing dynamic operations now benefit from substantial fidelity improvements thanks to Orbit, a new function from Quantum Elements achieving up to 27.51x gains on IBM Quantum hardware. This performance boost, demonstrated on IBM’s Heron processor, specifically addresses the challenges of maintaining coherence during circuits incorporating mid-circuit measurements and classical feedforward, a common feature in advanced quantum algorithms, the firm reports. Detailed results indicate Orbit’s impact extends beyond QFT circuits, with a 24x reduction in absolute error observed in Variational Quantum Eigensolver (VQE) calculations relative to ideal energy values.

Bernstein-Vazirani algorithms, used to find hidden bitstrings, also saw significant improvements; full Orbit workflows enabled successful evaluation of circuits reaching 70 qubits, exceeding the 26 qubits achieved with raw circuits and the 60 qubits of transpilation alone. Users can directly compare configurations with raw circuits to pinpoint which techniques yield measurable benefits for their specific circuit and backend, guiding more informed optimization decisions. This granular control is a key feature, allowing experimentation and fine-tuning of the optimization process.

Logical-Level Protection Boosts Bell-State Fidelity to 95.3%

Protecting entangled states from decoherence reached 95.3% fidelity using a new approach called Orbit, as demonstrated on IBM Quantum hardware. The gains extend beyond Bell states; Orbit demonstrably improved performance across a range of quantum algorithms and circuit types. This indicates that dynamic circuits are particularly susceptible to decoherence during idle periods, making them ideal candidates for dynamical decoupling.

Orbit Optimizes Variational Algorithms and Error Correction Workloads

Orbit extends its benefits beyond simple state preservation, demonstrably improving performance across diverse quantum algorithms and workloads. Experiments utilizing Variational Quantum Eigensolver (VQE) with an optimized unitary coupled-cluster ansatz to estimate the ground-state energy of LiH revealed Orbit’s capacity to reduce absolute error relative to ideal energy calculations; the improvement factor is calculated as the absolute difference between the measured and ideal energies with and without Orbit’s implementation.

The function’s utility extends to complex quantum-memory experiments employing heavy-hex and square-lattice surface codes, where encoded logical states are preserved over repeated rounds of error correction. It seamlessly integrates with fault-tolerant quantum computing, allowing combination with quantum error correction techniques, and supports both static and dynamic circuits, including those with mid-circuit measurements.

Stay current

See today’s quantum computing news on Quantum Zeitgeist for the latest breakthroughs in qubits, hardware, algorithms, and industry deals.

Dr. Donovan, Quantum Technology Futurist

Latest Posts by Dr. Donovan: