News

NVIDIA, IBM, and Fermilab Collaborate on Quantum-Classical Codesign Platform

NVIDIA, IBM, and Fermilab unveil a modular quantum computing platform and a 13,000x faster error-correction algorithm to advance quantum research.

NVIDIA, IBM, and Fermilab Collaborate on Quantum-Classical Codesign Platform

NVIDIA, IBM, and the U.S. Department of Energy’s Fermi National Accelerator Laboratory (Fermilab) have announced a collaboration to develop a modular quantum computing codesign platform aimed at advancing scalable, fault-tolerant quantum systems. This initiative integrates NVIDIA’s accelerated computing solutions with IBM’s quantum processors, including the IBM Quantum Nighthawk, the company’s most powerful quantum processor to date. Fermilab’s expertise in high-energy and nuclear physics (HENP) will further support the platform’s focus on quantum error correction and scientific applications.

Quantum-Classical Codesign Platform

The codesign platform will colocate NVIDIA DGX B200 systems with IBM quantum computers to enable researchers to develop and evaluate error-correction codes, decoding algorithms, and quantum-classical applications. This collaboration aims to address challenges in scaling quantum computers by exploring modular architectures that link multiple quantum processors. Fermilab’s contribution will focus on defining requirements for fault-tolerant quantum systems and advancing HENP research.

"Making quantum computers useful for science requires advances in error correction, applications, and the systems that bring quantum processors and accelerated computing together," said Krysta Svore, vice president of applied research for quantum computing at NVIDIA. Jerry Chow, IBM Fellow and chief technology officer of quantum-centric computing, added, "Scaling quantum and classical together will open new frontiers in high-energy and nuclear physics and beyond."

Breakthrough in Quantum Error-Correction Research

Separately, NVIDIA has introduced the TCBP-JOBS decoding algorithm as part of its CUDA-Q platform. This algorithm accelerates quantum error-correction research by up to 13,000 times compared to existing methods, reducing computation times from 12 weeks to just 45 minutes. The algorithm, available through the NVIDIA CUDA-Q QEC 0.9.0 library, enables researchers to evaluate error-correction codes more efficiently, paving the way for practical fault-tolerant quantum computing.

Quantum error correction is critical for mitigating errors inherent in quantum processors, which limit the complexity and duration of computations. The new decoding algorithm provides tools to optimize the use of physical qubits and improve performance measurements, significantly advancing the field.

Advancing Quantum Chemistry

In a related development, IBM and RIKEN have leveraged NVIDIA’s accelerated computing technologies to advance quantum-classical chemistry research. Their work, which combines IBM quantum computers with RIKEN’s GPU supercomputer ROQUO, has been recognized as a finalist for the 2026 ACM Gordon Bell Prize. The team successfully modeled protein-ligand complexes containing up to 12,635 atoms, a 40-fold increase over previous benchmarks, demonstrating the potential of quantum-classical approaches in drug discovery.

This collaboration between NVIDIA, IBM, and Fermilab represents a significant step toward integrating quantum and classical computing for scientific discovery, with applications spanning high-energy physics, quantum chemistry, and beyond.

Sign in to save this article Free account: build your research collection and return to read it anytime.