dorsaVi Confirms Suitability of RRAM Technology for Wearable AI-Driven Sensors
An internal assessment of dorsaVi’s (ASX: DVL) resistive random access memory (RRAM) technology has confirmed its suitability for high-speed, low-power usage in embedded, wearable and AI-driven devices.
Head-to-head comparisons with industry peers at the 40 nanometer (nm) node demonstrated the technology could offer highly-competitive speed, retention and endurance metrics.
High-speed performance with up to 200 nanosecond switching and ultra-low energy consumption establishes RRAM as a compelling option for edge computing demands.
Sensor Integration
dorsaVi’s expects integration of RRAM into its wearable sensors to significantly reduce latency, improve power efficiency and enable more autonomous decision-making at the edge.
RRAM’s dual binary and analog modes support memory and neuromorphic computing, opening future opportunities in robotics, adaptive sensing, and advanced edge-AI systems.
The technology will become a cornerstone of dorsaVi’s roadmap for embedded intelligence offering significant enhancements in sensor responsiveness, power efficiency and on-device computation for real-time biomechanics and movement monitoring.
Technology roadmap
dorsaVi chair Gernot Abl said the assessment supported the company’s roadmap toward delivering a new generation of ultra-efficient, AI-enabled sensor platforms.
“The integration of RRAM into edge-AI platforms is a pivotal step in our evolution toward intelligent, low-power motion systems,” he said.
“This technology enhances the performance of our existing solutions and and we are excited by the commercial and technical opportunities it creates as we continue to lead the industry in real-time movement analysis.”
Mr Abl said the company would continue internal and third-party testing to explore RRAM’s long-term stability, software integration and miniaturisation as it moves toward commercial deployment and strategic expansion into adjacent markets.
Next-gen memory
RRAM is a next-generation non-volatile memory technology that stores data by switching the resistance states of a metal-insulator-metal structure.
Unlike traditional memory, it operates in binary and analog modes to offer fast storage and retrieval, as well as tunable resistance for in-memory processing and synapse-like behaviour.
This dual capability is believed to position the technology at the forefront of conventional memory and adaptive learning convergence.
Mr Abl said RRAM was a “critical enabler” for applications where power efficiency and real-world adaptability are paramount such as medical-grade wearable sensors, Internet of Things, neuromorphic processors and advanced robotic systems.