IBM, Nvidia and Fermilab plan a co-design platform for modular quantum computers
IBM, NVIDIA and Fermi National Accelerator Laboratory announced plans on November 16, 2026 to build a co-design platform for modular quantum computers.
The platform centers on two flagship components: IBM's Quantum Nighthawk processor and NVIDIA's DGX B200 systems. IBM's Nighthawk features 120 qubits and can run circuits with up to 7,500 two-qubit operations, processing speeds reaching 100,000 samples per second.
NVIDIA's contribution comes from its DGX B200 accelerated computing systems, which have previously powered large AI workloads. Fermilab's SQMS Center will focus on connecting multiple quantum processors and scaling classical computing resources as the quantum system expands.
Error Correction Breakthrough
Alongside the announcement, NVIDIA revealed a new decoding algorithm called TCBP-JOBS designed for quantum error-correction research. The company claims the algorithm could reduce certain computation times from 12 weeks to just 45 minutes, representing a potential speedup of up to 13,000x.
The 13,000x figure is an upper bound, and the 12-weeks-to-45-minutes comparison applies to certain tasks.
Existing Partnerships
This collaboration extends established relationships. IBM's partnership with Fermilab's SQMS Center was formalized in 2024, while NVIDIA has been developing its CUDA-Q software for quantum applications, with the CUDA-Q Logical release arriving in September 2026.
IBM has also been active in quantum partnerships, including a November 2026 collaboration with Oak Ridge National Laboratory and AMD, another pairing of quantum hardware with a national laboratory and classical chipmaker.