Xanadu and Lockheed Martin Partner on Quantum ML
Quantum computing firm Xanadu and defense contractor Lockheed Martin have launched a joint research initiative to develop new hybrid quantum-classical algorithms. The partnership will focus on financial applications like portfolio optimization and risk modeling, while Xanadu continues to expand its software ecosystem by integrating its open-source PennyLane framework with more tools.
The collaboration will specifically target generative models, aiming to overcome the data-hungry and energy-intensive nature of classical AI. The research will explore Fourier-based operations, which are inaccessible to classical machine learning, to potentially create novel ways of representing and processing data in low-data environments common in finance and defense. This initiative is led by Xanadu's founder and CEO, Christian Weedbrook, and Dani Couger, the Quantum Technologies Lead for Lockheed Martin. The partnership aims to combine Xanadu's expertise in quantum-photonic hardware and software with Lockheed Martin's deep domain experience to redefine the foundations of how quantum computers learn. Xanadu builds its quantum computers using photonic technology, which operates at room temperature, a significant distinction from many competitors that require cryogenic setups. Their long-term goal is to build a fault-tolerant, one-million-qubit quantum data center by the end of the decade. This partnership is not Lockheed's only venture into quantum computing; they are also collaborating with PsiQuantum to develop applications for aerospace and defense. At the core of Xanadu's software strategy is PennyLane, a Python library for quantum differentiable programming that integrates with popular machine learning tools like PyTorch and TensorFlow. To enhance performance, Xanadu has integrated PennyLane and its Catalyst compiler with the Munich Quantum Toolkit (MQT), allowing users to access advanced compilation and optimization techniques with a single line of code. This focus on an accessible, interoperable software stack is designed to accelerate the development of hybrid quantum-classical applications.
