Can dorsaVi’s 22-nm RRAM Collaboration Deliver on Robotics Market Promise?
dorsaVi has launched a multi-year collaboration with leading Asian research institutes to develop its proprietary RRAM technology at the 22-nanometer node, targeting AI-driven robotics and wearable applications.
- Collaboration with Taiwan’s ITRI and Singapore’s NTU to advance RRAM technology
- Focus on 22-nm node for embedded non-volatile memory optimized for edge AI
- Target markets include robotics, wearables, electric vehicles, and industrial automation
- Neuromorphic computing integration to enable ultra-low-power, low-latency processing
- Robotics identified as primary near-term commercial opportunity
Strategic Collaboration to Push RRAM Technology
dorsaVi Limited (ASX, DVL) has announced the commencement of a significant multi-year collaboration with Taiwan’s Industrial Technology Research Institute (ITRI) and Singapore’s Nanyang Technological University (NTU). The partnership aims to advance dorsaVi’s proprietary Resistive Random Access Memory (RRAM) platform to the commercially critical 22-nanometer technology node. This development is a strategic inflection point for dorsaVi, positioning its RRAM technology to meet the demanding performance, power efficiency, and cost requirements of next-generation intelligent hardware.
The collaboration leverages dorsaVi’s deep intellectual property portfolio, NTU’s advanced device and materials research capabilities, and ITRI’s expertise in circuit co-design and back-end-of-line manufacturing processes. Together, they aim to deliver a scalable, reliable, and energy-efficient embedded memory solution compatible with TSMC’s 22-nm CMOS process, a key enabler for volume production.
Targeting High-Growth Markets with Edge AI Focus
The 22-nm RRAM platform is designed to support a broad range of high-growth markets, including robotics, wearables, electric vehicles, industrial automation, neuromorphic computing, and smart energy systems. dorsaVi has identified robotics as the lead commercial opportunity, driven by accelerating adoption of intelligent automation in industrial, medical, and service sectors.
By integrating compute-in-memory and neuromorphic processing capabilities, the platform addresses critical system requirements such as ultra-low power consumption, low latency, and deterministic real-time operation. This is particularly important for edge AI applications where fast, energy-efficient decision-making close to sensors is essential.
Neuromorphic Computing as a Foundation
dorsaVi’s approach incorporates neuromorphic computing principles, which mimic neural architectures to enable event-driven, always-on intelligence. This is a response to the limitations of conventional processor-centric designs, which face bottlenecks in power and memory bandwidth. The 22-nm RRAM platform’s multi-state memory and compute-in-memory macros provide the building blocks for efficient AI and robotics applications requiring adaptive, real-time processing.
The collaboration’s technical roadmap includes device and materials optimisation, circuit and peripheral co-design, and CMOS-compatible integration. Key performance targets include reducing write voltage and latency, improving endurance and retention, and achieving compute-in-memory efficiency exceeding 20 tera operations per second per watt (TOPS/W).
Commercial and Strategic Implications
While the financial impact of this collaboration is yet to be quantified, the advancement to the 22-nm node represents a compelling commercial scale-up opportunity for dorsaVi. The technology’s versatility and configurability allow it to be tailored to specific application needs, offering a competitive edge over generic memory solutions.
By focusing on robotics and edge AI, dorsaVi is positioning itself at the forefront of emerging markets that demand intelligent, low-power hardware. The partnership with globally recognised research institutions also enhances dorsaVi’s credibility and access to advanced semiconductor manufacturing ecosystems.
Bottom Line?
dorsaVi’s 22-nm RRAM collaboration sets the stage for next-gen AI hardware, but execution and market adoption remain key hurdles.
Questions in the middle?
- What is the expected timeline for prototype demonstration and commercial deployment?
- How will dorsaVi’s RRAM technology compete against established semiconductor memory players?
- What are the specific milestones and risk factors in scaling the 22-nm RRAM platform?
