Neuromorphic Quantum-Inspired Optimization and Simulation

forthcoming

NEQIOS introduces a radically new approach to solving complex computational problems that resist efficient solution on existing digital or quantum systems. We propose a compact neuromorphic platform that physically emulates key quantum properties, such as annealing, tunnelling, and superposition, without relying on fragile quantum states. Instead, NEQIOS integrates analogue in-memory computing (AIMC), integrated photonics (IPC), and reconfigurable nonlinear computing (RNC) in a high-performance system-on-package design. This innovation is urgently needed. From materials discovery to logistics and drug design, society faces NP-hard problems that scale exponentially. Quantum computers offer promise, but current systems – operating at cryogenic temperatures and requiring large infrastructure – remain costly, slow, difficult to scale and prohibitive for use on edge devices. NEQIOS offers a robust, room-temperature alternative that drastically reduces energy use (below 1 mW) and solution time (microseconds), while achieving solution quality on par with quantum annealers. NEQIOS’s key breakthrough lies in decomposing energy-based models into fine-grained computational primitives, each mapped to a dedicated physical substrate. AIMC performs high-throughput vector-matrix operations, IPC ensures precision and low-latency signal flow, and RNC handles nonlinear, adaptive function approximation. Together, these components form a physical solver capable of mimicking quantum behaviour with unmatched efficiency and scalability. NEQIOS is driven by a world-class European consortium with deep expertise in neuromorphic computing, photonic hardware, and analogue systems. With IBM Research Europe, Radboud University, University of Heidelberg, Enlightra, University of Oxford, and University of Twente, we unite technical excellence and a bold vision to deliver a transformative computing paradigm, aligned with Europe's semiconductor strategy and future demands in AI, science and industry.