Overview
To promote the application of quantum computing in the aerospace sector, Xanadu Quantum has formed a strategic alliance with Rolls-Royce and AMD. Using high-performance AMD GPUs and the PennyLane software framework, the team was able to effectively model the intricate computational fluid dynamics needed to develop contemporary jet engines. By carrying out a huge quantum algorithm on a scale previously believed to be unachievable in a quantum environment, this cooperation accomplished an important milestone. These innovations are aimed at producing more efficient airplanes by solving difficult linear equations faster than existing approaches. This article also emphasizes Xanadu’s overall expansion, including new financing, physical infrastructure, and its journey to become a publicly listed company.
Xanadu, Rolls-Royce, and AMD Scale Jet Engine Simulations on High-Performance GPUs
Today, Xanadu Quantum Technologies, Rolls-Royce, and AMD revealed the successful scaling of quantum algorithms intended to transform aerospace engineering in a historic partnership that heralds a new age for industrial quantum computing. The partners have carried out an unprecedented computational fluid dynamics (CFD) simulation by using AMD Instinct MI300X GPUs and Xanadu’s PennyLane software, bridging the gap between theoretical quantum research and real-world commercial applications.
The Problem: Going Beyond Traditional Supercomputing
The development of next-generation, low-emission jet engines is a huge undertaking for industry titans like Rolls-Royce. Computational Fluid Dynamics (CFD) is a major tool used in this procedure to simulate intricate airflow through engine turbines. Even on the most potent classical supercomputers in the world, these simulations now involve solving enormous systems of linear equations, which takes a very long time.
Classical computer power is turning into an unaffordable constraint as the industry shifts toward whole-engine simulations. In their joint announcement, the partners stated, “The challenge we face is mapping these massive problems to things that quantum computers are actually good at.” By using mathematical structure and interference to magnify accurate results and significantly reduce processing processes, quantum computing presents a potential solution.
You can also read Infleqtion Quantum Computing Expands DOE Labs Partnerships
The solution: The “Grand Unification” of Algorithms
The Quantum Singular Value Transformation (QSVT) method is the technological foundation of this innovation. QSVT, also known as the “grand unification” of quantum algorithms, allows an arbitrary matrix to be transformed using a given polynomial without explicitly computing its singular values.
Because QSVT needs a very small number of qubits (between 100 and 500) while retaining the depth required for complicated transformations, it is especially promising for first-generation fault-tolerant quantum computers (FTQC). For Rolls-Royce, this technique is the top candidate for solving the linear systems employed in CFD more scalably than ever before.
You can also read HBN Hexagonal Boron Nitride Defects Revealed by NV Sensors
35 million gates in two hours
By creating and running a 256 × 256 CFD matrix model, the partnership accomplished a historic milestone. Using 20 qubits and an astounding 35 million quantum gates, the simulation was finished in less than two hours on a single AMD Instinct MI300X GPU. This is a difficulty scale that has never been used in a quantum environment previously.
The inclusion of Xanadu’s Catalyst, a just-in-time (JIT) compiler, which enabled the team to build the program once and run it with many settings at runtime, was crucial to this achievement. “Seeing AMD high-performance compute boost the performance of PennyLane is a clear proof point of how quantum and classical technologies can effectively work together,” stated Madhu Rangarajan, Corporate Vice President at AMD. This hardware-software synergy led to a 25x simulation speedup on AMD GPUs compared to traditional multithreaded CPU execution.
You can also read Infleqtion At Analyst Day 2026 Highlights Quantum Roadmap
Bridging the Software Gap
Rolls-Royce started the procedure by creating meshes that represented a 2D lid-driven cavity test case. Large, sparse, non-Hermitian matrices were produced by linearizing non-linear fluid equations after these meshes were transformed into square matrices.
The group carried out two crucial actions using PennyLane:
- Embedding: Using custom methods that capitalized on the sparse structure of the matrix, the matrix was effectively encoded into the quantum system.
- Transformation: A GPU-compatible version of the Non-linear Fast Fourier Transform was created in PennyLane to swiftly transform inverse polynomial approximations into quantum operations.
To ensure that the software stack is “ready to fly” by the time bigger fault-tolerant hardware arrives, this enables researchers to formulate complicated algorithms and compile them for high-performance execution.
You can also read BTQ NASDAQ Selected for Australia’s Industry Growth Program
Xanadu’s New Public Era
For Xanadu, the news is made at a critical juncture. The Toronto-based firm and Crane Harbor Acquisition Corp. recently announced a business combination agreement. With a projected gross proceeds of US$500 million, the merged company, Xanadu Quantum Technologies Limited, is anticipated to list on the Nasdaq and Toronto Stock Exchanges. With this action, Xanadu is anticipated to become the first and only “pure-play” photonic quantum computing business to be publicly listed.
The Path Forward
The partners emphasized that quantum simulation is still essential for verifying research and setting up the industrial software ecosystem, even though full-scale fault-tolerant quantum hardware is still a ways off. “The path to quantum advantage is not a solo journey,” stated Xanadu’s founder and CEO, Christian Weedbrook. To move quantum computing from research settings to practical engineering applications, partners like Rolls-Royce and AMD’s hardware leadership are necessary.
The capacity to assemble and optimize programs of this size will be crucial for the design of the next generation of effective, low-emission aircraft as aerospace companies seek to get a competitive edge.
You can also read Quantinuum at the APS Global Physics Summit 2026 in Denver
