Nvidia has released what it says is the world's first family of open-source artificial intelligence models designed specifically for quantum computing, embedding its GPU and software stack deeper into the emerging field.
The Ising model family, named after the mathematical framework used to simplify complex physical systems, targets two critical bottlenecks standing between today's fragile quantum hardware and fault-tolerant machines: processor calibration and error correction.
Ising Calibration is a 35 billion-parameter vision-language model trained to read experimental data from a quantum processing unit (QPU) and infer the adjustments needed to tune it, reducing setup time from days to hours.
The Ising Decoding models, a pair of compact neural networks optimised for speed and accuracy, deliver quantum error correction that Nvidia says is up to 2.5 times faster and three times more accurate than the open-source decoder most research groups currently use.
Early adopters include Harvard, Fermilab, Lawrence Berkeley National Laboratory, the UK's National Physical Laboratory and quantum hardware makers IQM and Infleqtion.
While the models themselves are open-source, the infrastructure they run on is not: the decoder relies on Nvidia's NVQLink interconnect and CUDA-Q software platform, tying quantum researchers more tightly into the company's hardware ecosystem.
The launch coincided with an expanded partnership with Cadence Design Systems, the electronic design automation (EDA) company, announced at CadenceLIVE Silicon Valley this week.
The collaboration integrates Cadence's chip design and system analysis tools with Nvidia's CUDA-X libraries and Omniverse simulation platform, with the two companies claiming potential speedups of up to 100 times across engineering workflows.
Samsung, SK Hynix and Argonne National Laboratory are among those already using the accelerated tools.
Together, the moves extend Nvidia's reach beyond its core GPU business into two adjacent markets: the control layer of quantum computing and the design workflows used to build the next generation of semiconductors.
Nvidia shares have risen roughly 18% over the past ten trading sessions and were trading around $200 on Thursday, about 8% below their all-time high.
Analysts remain broadly bullish, with a median price target of $265 among 70 covering firms, implying roughly 33% upside from current levels.
The recap
- NVIDIA launches Ising open-source quantum AI model family.
- Simply Wall St projects $559.7 billion revenue by 2029.
- Cadence partnership expands NVIDIA’s role in chip design workflows.
