NVIDIA, IBM Build Modular Quantum Computing Platform
The codesign platform pairs IBM quantum computers with NVIDIA DGX B200 systems for error-correction and hybrid research.
Updated on Nov. 21, 2026 in Quantum Computing

NVIDIA, IBM and Fermilab, a U.S. Department of Energy laboratory, are developing a modular quantum computing codesign platform combining IBM quantum computers with NVIDIA DGX B200 systems. NVIDIA also announced the TCBP-JOBS decoding algorithm for quantum error-correction research.
Why it matters
Quantum processors are vulnerable to errors that limit the complexity and duration of calculations, and faster error decoding could let researchers evaluate correction methods more efficiently. The platform aims to connect quantum and classical hardware within a single research workflow.
NVIDIA says the TCBP-JOBS algorithm, available through the CUDA-Q QEC 0.9.0 library, is up to 13,000 times faster than leading alternatives, and the claim comes from NVIDIA rather than independent testing. The collaboration also expands Slurm workload-manager support so quantum processors can be scheduled alongside CPUs and GPUs.
The players
NVIDIA
NVIDIA is a U.S. semiconductor and accelerated computing company known for its GPUs and the CUDA software platform.
IBM
IBM is a multinational technology company that operates one of the largest quantum computing research programs in the world.
Fermilab
Fermilab is a U.S. Department of Energy national laboratory near Chicago focused on particle physics and related computing research.
The details
The platform will combine IBM quantum computers with NVIDIA DGX B200 systems to support quantum error correction, scientific applications and hybrid quantum-classical computing. IBM will contribute integration through its Quantum Resource Management Interface, and IBM and Japan's Riken research institute's chemistry work modeling protein-ligand complexes was named a finalist for the 2026 ACM Gordon Bell Prize.
Timeline
Nov. 21, 2026: Core platform event date reported.
2026: ACM Gordon Bell Prize finalist recognition for the Riken-IBM chemistry research.
The Tech Race
The TCBP-JOBS decoder extends NVIDIA's CUDA-Q QEC library, placing NVIDIA's accelerated-computing stack at the center of the race to make quantum machines usable alongside classical supercomputers. The platform also positions IBM's quantum hardware within a GPU-driven hybrid architecture.
For most people, the immediate effect is indirect: faster error-correction research could shorten the path to quantum machines that solve chemistry and materials problems. Researchers at universities and national labs are the most direct users of platforms like this one.
The takeaway
Error correction remains the bottleneck for useful quantum computing, so faster decoders matter as much as qubit counts. Watch for independent benchmarks that confirm vendor speedup claims before relying on them for research planning.
Further reading
For background on the hardware driving these efforts, see our Quantum Computing coverage.
Source note: This article includes information reported by The Manila times.



