EMASS, a subsidiary of Nanoveu, has announced a major upgrade to its edge AI chips. The new ECS-DoT SoC built on a 16nm process is nearing the final stages of development. It is heading toward design sign-off, tape-out, and manufacturing at TSMC. This new chip builds on the existing 22nm ECS-DoT, offering more power and performance for always-on AI applications.

Next-Generation Edge AI with Enhanced Capabilities

The 16nm ECS-DoT enhances EMASS’s lineup by delivering better energy efficiency, greater compute power, and expanded on-chip resources. It maintains the core architecture, developer workflow, and ultra-low-power design principles of the 22nm version. The chip includes an integrated Bluetooth Low Energy (BLE) subsystem, removing the need for external wireless modules. This reduces board size, costs, and design complexity.

Additional improvements include a larger on-chip SRAM and a new power management system. These enhancements allow the chip to handle more demanding AI models and higher workloads while consuming less energy. The integration of an object-detection accelerator helps process vision tasks faster, reducing latency and increasing throughput for edge vision applications. The floating-point unit (FP16/FP32) accelerates digital signal processing and mixed-precision AI workflows, making development easier.

Ease of Development and Broader Applications

The new ECS-DoT maintains full compatibility with EMASS’s existing software and development workflows. Developers can upgrade from the 22nm to the 16nm version with minimal redesign effort. This means they can deploy larger, more complex AI workloads without sacrificing power efficiency or increasing complexity.

By migrating to 16nm, EMASS aims to unlock new use cases that demand more intelligence, faster processing, and even lower power consumption. The chip is designed for applications like smart sensors, IoT devices, and other battery-powered systems that need to stay always-on. The upgrade supports EMASS’s vision of integrating AI from the smallest components to full applications.

Overall, the 16nm ECS-DoT marks a significant step forward for EMASS. It pushes the boundaries of edge AI technology, enabling smarter, more capable devices that can operate longer on a single charge. The company’s focus on high integration, energy efficiency, and developer support aims to keep it at the forefront of ultra-low-power edge AI solutions.

Inspired by

• https://ai-techpark.com/emass-pushes-the-edge-further-with-16nm-ecs-dot/