Designed for edge inference applications, the Origin E6 NPU offers a state-of-the-art solution supporting both traditional and generative AI models. With performance scaling from 16 to 32 TOPS, the E6 is tailored to meet the sophisticated demands of next-gen smartphones, AR/VR devices, and automotive applications. Expedera's packet-based architecture facilitates parallel operations, ensuring efficient resource use and reducing latency.
The E6 is highly customizable, supporting a wide range of AI models and enabling seamless deployments across diverse hardware environments without the need for extensive hardware-specific tuning. Expedera offers the Origin E6 with a complete software stack that promotes ease of integration and supports advanced features like multiple job APIs and quantization options.
By achieving up to 90% processor utilization, the E6 maximizes performance while minimizing dark silicon waste, positioning it as a leading choice for AI chip architects seeking power efficiency and scalability in their designs.
