QpiAI cuts quantum error correction time to 1.5 microseconds

QpiAI has introduced decoder hardware for quantum error correction on its 64-qubit Kaveri superconducting processor, cutting correction time to about 1.5 microseconds per cycle.

The decoder targets one of quantum computing's central engineering challenges: identifying and correcting errors fast enough to keep calculations stable. Conventional software methods can take tens of microseconds for the same task, while QpiAI's hardware-based approach completes error detection and correction within the coherence window of superconducting qubits.

The Bangalore-based company's platform uses a union-find algorithm and operates with a distance-5 rotated surface code across 49 physical qubits. Each decoder instance runs on a single Kaveri quantum processing unit, allowing one decoder per chip and supporting integration with existing quantum hardware.

The work falls within India's National Quantum Mission, which has provided partial funding as part of the country's broader push to build domestic quantum computing and related technologies.

The current platform supports up to 20 decoders running in parallel, enabling simultaneous error correction across multiple logical qubits. It performs five rounds of stabiliser measurements per cycle to detect both qubit errors and measurement errors.

The decoder completes distance-5 surface code decoding in up to 40 clock cycles. According to technical details released by QpiAI, the superconducting platform operates with approximate T1 times of 100 microseconds and T2 times of 95 microseconds, leaving room for multiple correction cycles before decoherence affects computation.

Hardware Shift

Quantum error correction has long been seen as essential to building larger, more reliable quantum systems, but it typically relies on classical computing resources to process the error data generated by qubits. By moving decoding into dedicated hardware, QpiAI is aiming to reduce latency and lessen reliance on external CPUs and GPUs.

Nagendra Nagaraja, founder and chief executive officer of QpiAI, said, ""The performance of our new decoder platform demonstrates a practical pathway toward scalable, hardware-accelerated quantum error correction."

He added: "Compatible with widely used superconducting transmon qubits, the platform limits the need for additional classical support from CPUs and GPUs. QpiAI is also developing next-generation error correction methods tailored to our own fluxonium-based qubits as well as architectures designed to operate across distributed systems."

QpiAI said its processors combine error-correction architecture with superconducting processor layout and fabrication choices intended to support this approach. The decoder also works with the company's QpiAI Sense quantum control electronics, which are being used in the development of larger-scale systems.

National Backing

The project has also drawn support from Indian government officials involved in the National Quantum Mission. Dr Abhay Karandikar, secretary of the Department of Science and Technology, described the result as an important milestone in the programme.

"Quantum error correction (QEC) is essential for scalable quantum computing," Karandikar said. "By implementing distance-5 surface code QEC in custom hardware rather than traditional CPUs, QpiAI is accelerating the deployment of its 64-qubit Kaveri QPU in India, marking a major step toward practical, large-scale quantum utility."

The current system has been validated using simulated qubits and is now undergoing integration and experimental validation on physical qubits. That means the decoder's reported performance has so far been established in a controlled setting, with further testing under way on live superconducting hardware.

QpiAI's Kaveri processor is designed with qubit connectivity to support surface code error correction, widely seen as one of the most practical ways to manage errors in superconducting quantum systems. The decoder supports both Pauli errors and measurement errors, common features of noisy quantum environments.

For India's quantum sector, the work reflects a push to show progress not only in processor design but also in the supporting systems needed to make quantum machines usable at larger scales. QpiAI said the decoder enables active, closed-loop correction, allowing qubit fixes to be applied during execution and reducing the build-up of errors over time.