Low Power Neural Network Processor
Power and Area Optimized AI-Micro Core for Resource-Constrained IoT Sensors
OVERVIEW
The LP-NNP is a power and area-optimized Neural Network Processor designed for resource-constrained IoT sensors. It is an AI micro-engine designed to enable efficient neural network processing on low-power IoT sensors. The size of the engine is configurable for various applications. It comes with a comprehensive firmware that adapts existing machine learning learning libraries to its functionalities
Benefits
- Power optimized
- Area optimized
- Configurable
- AMBA Standard Interface
- AXI/AHB/APB
- UVM verification environment
- FPGA Validation Platform
- Comprehensive Firmware
APPLICATIONS
Industrial IoT
Smart Grid
Smart Cities
Low Power Neural Network Processor
The Low Power Neural Network Processor is designed to enable distributed learning on IoT. It is fully customizable comes with a firmware that adapts functionalities from existing ML libraries to the engine implementation.
