New Platform for Quantum Error Correction

IBM, in collaboration with the US Department of Energy-funded Quantum Science Center (QSC) at Oak Ridge National Laboratory (ORNL) and AMD, has introduced a co-design platform for quantum error correction that integrates quantum and high-performance computing. Over recent years, investigators have achieved significant progress in quantum error correction, including preliminary demonstrations of protecting quantum data from environmental noise, attaining quantum advantage over classical methodologies with verified accuracy, and proving that integrated quantum-classical systems can address industry-relevant challenges. The current research imperative involves scaling these successes through the utilization of advanced classical computing resources.

To meet this objective, IBM has initiated a co-design platform for quantum error correction with the QSC and AMD serving as founding partners. The initial implementation co-locates IBM’s quantum hardware, featuring its sophisticated 120-qubit Nighthawk processor, with ORNL’s Umbra High-Performance Computing (HPC) system, which utilizes AMD Instinct MI355X GPUs and 5th Gen EPYC CPUs. This AMD-driven infrastructure provides the classical computing base required for research into quantum error correction, decoding optimization, and complex software orchestration.

This co-design environment will enable researchers to analyze the deep integration of classical and quantum systems to sustain progress in error correction research. Such initiatives are a fundamental element of the Department of Energy’s Quantum Genesis Mission, which, via its Q Competition, seeks to achieve 100 logical qubits executing 100,000 operations by late 2028. Recent progress by IBM has demonstrated experiments utilizing up to 70 logical qubits and 438 operations. Bridging the remaining disparity necessitates approximately a 230-fold increase in the count of hard operations.

The Nighthawk processor’s lattice connectivity, circuit execution speed, and real-time pulse control enable researchers to evaluate various error correction methodologies. It allows teams to assess the specific requirements of each method regarding qubits, time, and classical resources to identify where GPU accelerators, such as those within the Umbra system, provide the most significant benefit. This platform will link with the Oak Ridge Leadership Computing Facility (OLCF) in Tennessee, establishing a path to explore quantum technology as a component of broader AI and scientific computing infrastructure. This ensures that quantum, AI, and high-performance computing evolve as an integrated ecosystem.

IBM’s Quantum System API provides investigators with sophisticated, pulse-level control over quantum hardware. On these systems, researchers can measure qubits during calculations and respond to findings in real time, calibrate the precise timing of signals protecting qubits from noise, verify hardware execution, and interact directly with raw measurement data. Developing the next generation of error correction approaches requires these advanced capabilities to operate in tandem on systems of sufficient scale to evaluate functional codes.

Research conducted via this co-design platform supports the Quantum Genesis Mission by developing error correction protocols that accelerate the delivery of scientifically relevant quantum computing by 2028. Attaining this objective requires workflows that pair quantum processors with error correction techniques supported by high-performance classical systems. These workflows encompass concatenated and space-time codes, algorithm-aware decoders, and hardware-aware circuit optimization, all integrated within a single technological stack.