• Exclusive and transformational neuromorphic portfolio secured from Technion led by Prof. Shahar Kvatinsky
• Processing-in-Memory (PIM) merges computation and memory, bringing brain-like reflexes, learning, and calibration directly into hardware
• Dual-layer neuromorphic design combines 'Reflex Engine' and 'Adaptive Interface' for autonomous, noise-resilient sensing and actuation

dorsaVi Limited (ASX: DVL) has announced the acquisition of a breakthrough neuromorphic enabling technology portfolio that elevates its biomedical and robotics reflex platforms into a new era of adaptive, next-generation intelligence. The acquired portfolio is pioneered by Prof. Shahar Kvatinsky, one of the world's leading experts in Processing-in-Memory (PIM) and neuromorphic hardware. PIM merges computation and memory into one intelligent substrate, bringing brain-like reflexes, learning, and calibration directly into hardware. The portfolio combines a 'Reflex Engine' for in-memory inference and on-device calibration, and an 'Adaptive Interface Layer' for noise-resilient signal conditioning and real-time actuation. This architecture is intended to support event-driven processing, millisecond-class response, and adaptive thresholds for tighter closed-loop behaviour and improved energy efficiency on dorsaVi's FDA-cleared biosensor and robotic platforms. The acquisition positions dorsaVi to participate in the rapidly growing neuromorphic computing market, which is forecast to expand from ~US$5.3 billion in 2023 to over US$20 billion by 2030, as the industry transitions from reactive analytics to self-adapting edge systems.

The acquired neuromorphic portfolio is expected to enhance on-device intelligence in dorsaVi's platforms by providing cleaner sensor inputs, faster and more efficient control loops, and improved application outcomes in areas like biosensing, rehabilitation, industrial ergonomics, and robotics. The technology enables wearables to generate real-time personalized therapy prompts, adaptive haptic patterns, and clinician-ready progress notes without streaming raw data, while neuromorphic controllers in prosthetics and robotics can generate intent-aligned motion profiles that adapt to changing contexts.