• Sub-microseconds array latency per reflex decision
• Picojoule (pJ) array conduction efficiency
• Binary writes < 200 pJ/bit supports nanojoule (nJ) typical event logs

dorsaVi Ltd (ASX:DVL) has announced a major technical milestone, with the successful engineering validation of its RRAM-powered Reflex Platform. The validation confirms the platform's potential to deliver biological-grade reflexes at unprecedented speed and energy efficiency. The key findings include: sub-microsecond array latency per reflex decision, picojoule (pJ) array conduction efficiency, and binary writes under 200 pJ/bit supporting nanojoule (nJ) typical event logs. These performance metrics enable the Reflex Platform to move from concept to deployment in robotic grippers, prosthetic limbs, exosuits, and next-generation human-machine interfaces, where speed, adaptability, and energy efficiency are paramount. The validation also reinforces dorsaVi's strategic direction and positions the company as a future leader in neuromorphic robotics, adaptive motion platforms, and embedded biosignal AI. The results further confirm that RRAM is no longer the limiting factor in achieving biological-grade sub-millisecond reflexes, with the performance ceiling now shifting to external components in the signal chain such as sensors, signal converters, and mechanical actuators.

These validation findings confirm that dorsaVi's RRAM-enabled Reflex Platform is ready for integration into real-world robotic and biomedical systems, unlocking capabilities that were previously unattainable with conventional memory architectures. Key application domains include reflex robotics, prosthetics and exosuits, and human-machine interfaces, where the platform can enable biological-like slip reflexes, EMG-responsive limb adaptation, and always-armed, instantly responsive wearable devices.