Key Highlights
- MosChip collaborates with EMASS on ECS-DoT Edge AI SoC development.
- The SoC targets up to 93 percent faster processing performance.
- Energy use reduced by up to 90 percent over conventional edge AI chips.
- Supports 22nm silicon implementation including physical design and packaging.
- Designed for wearables, drones, healthcare trackers and industrial IoT.
MosChip Technologies announced its collaboration with EMASS to support the silicon implementation of the ECS-DoT Edge AI System-on-Chip, an ultra-low-power chip designed for always-on intelligence across wearables, drones, industrial IoT systems, and edge sensors.
EMASS led the architecture design of the ECS-DoT SoC, targeting up to 93% faster processing and nearly 90% lower energy consumption than conventional edge AI solutions. MosChip contributed engineering support across 22nm silicon implementation, including physical design workflows, tape-out coordination, packaging, assembly, evaluation hardware, and validation activities.
The ECS-DoT SoC is based on a RISC-V architecture and integrates dual neural accelerators and 4MB of on-chip memory. It enables real-time, milliwatt-class inference for vision, audio, and sensor workloads without cloud dependency. Its compact, power-efficient design makes it suitable for wearables, healthcare devices, drones, and industrial monitoring systems.
Both companies noted that the partnership combines EMASS’s differentiated architecture with MosChip’s engineering expertise, resulting in a fully functional and robust SoC for OEM partners.