April 14th news, NVIDIA officially announced the launch of the world's first open-source quantum artificial intelligence model series - "Ising (Ising)". This series of models aims to tackle the core engineering bottlenecks in quantum computing, helping researchers and companies build fault-tolerant quantum processors that can run practical applications. Driven by this major positive news, the quantum technology sector rose collectively on the morning of April 15th, with Shenzhou Information hitting a limit-up, while China Xidian, Guodun Quantum, Luobo Tech, and other related stocks led the rise, with Luobo Tech seeing net inflow of nearly 500 million yuan from main fund flows.
The Ising series consists of two core models: Ising Calibration (Calibration) and Ising Decoding (Decoding). As a vision-language model, Ising Calibration can automate the calibration tasks of QPU (quantum processor), addressing hardware instability and parameter drift; Ising Decoding uses 3D convolutional neural networks for quantum error correction decoding, with a computing speed 2.5 times faster than the current industry standard and an accuracy improvement of up to 300%. Jensen Huang, CEO of NVIDIA, pointed out that Ising will make AI the "control plane" and "operating system" of quantum machines, transforming fragile qubits into scalable and reliable "quantum-GPU systems".
Currently, the model has been open-sourced on GitHub and Hugging Face, and has been adopted by research institutions such as Harvard University and Berkeley Lab. With the quantum computing market expected to exceed 11 billion US dollars by 2030, NVIDIA's move not only solidifies its ecological leadership in the high-performance computing field, but also marks the transition of quantum computing from laboratory environments to AI-driven fully automated operations, accelerating the commercialization of general-purpose quantum computers.
