Nvidia Releases Open-Source Quantum AI Model Ising, Quantum Computing Sector Rises Collectively, IonQ Gains Over 20%

TradingKey - On April 14, 2026 local time, NVIDIA ( NVDA) officially released the open-source quantum AI model NVIDIA Ising under the Apache-2.0 commercial license. The model focuses on two core engineering challenges: quantum processor calibration and quantum error correction.

Following the announcement, U.S. quantum computing stocks surged across the board: XNDU ( XNDU) rose 29.07%, SEALSQ ( LAES) jumped 21.03%, IONQ ( IONQ) gained 20.16%, D-Wave Quantum ( QBTS) climbed 15.84%, Quantum Computing Inc. ( QUBT) added 11.55%, Rigetti Computing ( RGTI) increased 11.50%. Among them, IONQ's single-day gain reached its highest level since late February.

What is the origin of the name Ising?

Ising is named after physicist Ernst Ising. In 1920, physicist Wilhelm Lenz proposed a model describing the interactions of particles in a lattice, and in 1925, his student Ernst Ising completed research on the one-dimensional case. Known as the Ising model, it is one of the classic models of statistical physics. Its core concept is that simple microscopic interactions can produce macroscopic phase transitions and collective behavior; it was originally used to explain the phenomena of magnets demagnetizing when heated and remagnetizing when cooled.

The principles of the Ising model have been widely applied in AI, financial markets, public opinion propagation, and quantum computing. Nvidia has named its quantum AI model 'Ising,' aiming to deeply integrate the quantum computing control layer with its own GPU computing ecosystem to build a technical architecture of 'AI Control System (Ising) + GPU Computing Platform (CUDA-Q).'

What problems does the Ising model solve?

Quantum computing faces two core hurdles: slow calibration and high error rates. Currently, the most advanced quantum processors can fail once every 1,000 operations, whereas truly functional quantum computers require error rates below one in a trillion. The Ising model is designed specifically to address these two challenges.

NVIDIA Founder and CEO Jensen Huang stated: "AI is critical to achieving practical quantum computing. With Ising, AI will serve as the operating system for quantum computers, transforming fragile qubits into scalable and reliable quantum-GPU systems."

Ising consists of two components.

Ising Calibration Is a 35-billion-parameter vision-language model used for automating the calibration of quantum processors. It can rapidly interpret quantum processor measurement results and drive AI agents to perform continuous proactive calibration. According to NVIDIA data, the model can reduce calibration time from days to hours. Training data covers multiple technical paths, including superconducting qubits, ion traps, neutral atoms, quantum dots, and electrons on helium.

In the QcalEval benchmark developed jointly by NVIDIA, Fermilab, Harvard University, and other institutions, Ising Calibration outperformed closed-source models such as Gemini 3.1 Pro, GPT 5.4, and Claude Opus 4.6 across six evaluation dimensions, including interpreting experimental results, classifying results, assessing result significance, evaluating fit quality and key features, and generating feasibility recommendations.

Ising Decoding Is an error-correction decoding model based on 3D convolutional neural networks, specifically designed for real-time decoding in quantum error correction. Users only need to define the noise model, the orientation of the rotated surface code, and the model depth, and the framework can automatically generate synthetic data to train an optimized 3D CNN. The model is available in two versions: the speed version has 912,000 parameters and is 2.5 times faster with 1.11 times higher accuracy than the open-source pyMatching solution; the precision version has 1.79 million parameters and is 2.25 times faster with 1.53 times higher accuracy.

Multiple leading institutions have integrated for testing.

Upon its initial release, the Ising model was adopted by several leading international research institutions and quantum enterprises. Ising Calibration has been deployed at Fermilab, Lawrence Berkeley National Laboratory, the National Physical Laboratory in the UK, Harvard University, IONQ, and Infleqtion. Ising Decoding has been put to the test at Sandia National Laboratories, Cornell University, the University of Chicago, and IQM Quantum Computers. NVIDIA simultaneously open-sourced all model code and released operational guidelines and NIM microservices, supporting local deployment to ensure the security of proprietary data.

IonQ surges over 20%

IONQ shares surged 20.16% for the day. For the full year 2025, IONQ reported revenue of $130 million, up 202% year-over-year; its 2026 revenue guidance of $225 million to $245 million represents year-over-year growth of approximately 81%. As of the end of 2025, its cash balance was approximately $3.3 billion. IONQ has fallen about 20% year-to-date, and this rebound has partially recovered its earlier losses.

Market Perspectives

Research firm Resonance predicts that the global quantum computing market will exceed $11 billion by 2030. Nvidia's entry is viewed by the market as a signal of the industry's accelerated evolution from proof-of-concept to engineering implementation.

In terms of ratings, IONQ has received 9 "Buy" and 3 "Hold" ratings over the past three months, resulting in a consensus rating of "Strong Buy"; the average price target of $65.91 implies approximately 84% upside from current levels.

From a risk perspective, IONQ's 2025 GAAP net loss was $510.4 million, with an adjusted EBITDA loss of $186.8 million, and the EBITDA loss is projected to widen to between $310 million and $330 million in 2026. Quantum computing companies overall remain in the early stages of commercialization with unestablished profitability, and the risk of short-term stock price volatility should not be overlooked.

In summary, the sector's rally is a direct response to Nvidia's technical capabilities. The subsequent trend depends on the scalability verification of the Ising model in real-world deployment scenarios and whether industry leaders like IONQ can continue to deliver on high-growth expectations. The sector currently remains in a phase where theme-driven momentum and fundamental verification proceed in parallel.