Nvidia unveiled Ising, an open-source family of AI models designed to calibrate quantum processors and decode errors—two sticking points on the road to practical quantum computing. The company says the tools deliver real-time decoding up to 2.5 times faster and three times more accurate than the open-source pyMatching baseline, while a vision-language model automates calibration, cutting workflows from days to hours. Ising plugs into Nvidia’s CUDA-Q software stack and NVQLink interconnect for tight GPU–QPU control and can be fine-tuned via NVIDIA NIM microservices or run locally to keep proprietary data in-house. Early adopters include Atom Computing, IonQ, IQM, U.S. national labs and leading universities. CEO Jensen Huang framed AI as the “control plane” for quantum machines, underscoring Nvidia’s bid to anchor hybrid quantum–classical systems on its GPU platform. Analyst firm Resonance pegs the quantum market at more than $11 billion by 2030, though Nvidia cautioned timelines and features remain subject to change.
NVIDIA Debuts Ising, the First Open Quantum-AI Model Suite Aimed at Making Quantum Computers Practical
- Trending
- Comments
- Latest
