The DesignWare ARC EV6x Embedded Vision Processors offer an optional convolutional neural network (CNN) engine for fast and accurate object detection, classification and scene segmentation. The EV6x supports any CNN, including popular networks such as AlexNet, VGG16, GoogLeNet, Yolo, Faster R-CNN, SqueezeNet, and ResNet.
Designers can run CNN graphs originally trained for 32-bit floating point hardware on the EV6x’s 12-bit CNN engine, significantly reducing the power and area of their designs while maintaining the same levels of detection accuracy as a 16-bit engine. The EV6x’s CNN hardware also supports 8-bit precision for lower memory bandwidth and power use.
The EV6x’s CNN engine delivers power efficiency of up to 2,000 GMACs/sec/W when implemented in 16-nm FinFET process technologies (worst-case operating conditions).
The MetaWare EV Development Toolkit includes a CNN mapping tool that analyzes neural networks trained using popular frameworks like Caffe and Tensorflow, and automatically generates the executable for the programmable CNN engine. For maximum flexibility and future-proofing, the tool can also distribute computations between the vision CPU and CNN resources to support new and emerging neural network algorithms as well as customer-specific CNN layers.
Inuitive Demonstrates NU4000 Artificial Intelligence SoC with DesignWare EV6x at CES 2019
Downloads and Documentation
Fast, accurate object detection with a programmable CNN engine
Offers power efficiency of up to 2000 GMACs/sec/W in worst-case operating conditions when implemented in 16-nm FinFET processes
Supports both coefficient and feature map compression/decompression to reduce data bandwidth requirements and decrease power consumption
High productivity MetaWare EV Development Toolkit’s CNN mapping tool automatically dispatches processing tasks to available hardware resources for faster execution