MCUs accelerate multimodal edge AI
The Synaptics Astra SR200 series of xMCUs runs multimodal AI workloads without the memory and compute limitations of conventional MCUs. Devices combine heterogeneous, AI-native compute and acceleration with integrated high-bandwidth custom DRAM scalable to hundreds of Mbytes. This allows larger AI models and richer multimodal workloads to run entirely on-device.
Powered by Synaptics’ Torq AI engine with Google Research’s transformer-capable Coral NPU, the xMCUs deliver up to 400 GOPS of AI inference. They also feature Arm Cortex-M85 and Cortex-M52 cores, integrated vision and audio capabilities, and hardware-rooted security. A tiered architecture pairs up to 8 Mbytes of SRAM with extensible, high-bandwidth, high-capacity custom DRAM. Synaptics says this architecture runs workloads at one-third the power of typical system DRAM.
The Astra SR200 series comprises three pin-compatible product lines sharing a common hardware foundation and SDK. The SR210 reaches 200 GOPS for always-on voice and sensing; the SR230 adds vision, display, and optional audio at 300 GOPS; and the SR250 provides 400 GOPS with dual vision and display paths for demanding edge applications.
Locate a distributor or sales office here.
The post MCUs accelerate multimodal edge AI appeared first on EDN.
