NVIDIA wants to be the operating system for quantum computers. The company just released Ising, a family of open-source AI models that tackle the two biggest headaches in quantum computing: calibrating processors and correcting errors. The timing matters. Quantum computing is supposed to be an $11 billion market by 2030, according to analyst firm Resonance, but only if someone solves these engineering problems first.
The Ising suite includes two main tools. Ising Calibration is a vision language model that reads measurements from quantum processors and automates continuous calibration, cutting what used to take days down to hours. Ising Decoding uses 3D convolutional neural networks to handle quantum error correction in real-time, with NVIDIA claiming it is up to 2.5x faster and 3x more accurate than pyMatching, the current open-source standard. Both models run locally on researchers' systems, which matters when you're dealing with proprietary hardware data.
This move is rough for quantum software startups. NVIDIA is essentially commoditizing the middleware layer that many of these companies built their businesses around. When a hardware giant with NVIDIA's resources gives away tools for calibration and error correction, startups selling proprietary versions of the same thing need to find new ground to stand on. This dynamic mirrors the ongoing software battles in the broader AI ecosystem, where competitors are attempting to build alternative stacks to AMD's ROCm as a CUDA alternative. The models plug into NVIDIA's CUDA-Q platform, integrate with their NVQLink hardware interconnect, and are available now on GitHub, Hugging Face, and build.nvidia.com. Academic labs and quantum companies including IonQ, Infleqtion, and IQM Quantum Computers are already adopting the tools.
Jensen Huang called AI "the control plane" for quantum machines. By open-sourcing these models and tying them to the existing CUDA ecosystem, NVIDIA is making itself essential infrastructure for quantum computing before the industry has even figured out what useful applications look like. Some competitors, however, view the dominance differently: D-Wave CEO claims NVIDIA should be 'shaking in their boots'.