DesignWare EV7x Vision Processors

Delivers Industry-Leading 35 TOPS Performance for Artificial Intelligence SoCs

The DesignWare EV7x Vision Processors’ heterogeneous architecture integrates vector DSP, vector FPU, and a neural network accelerator to provide a scalable solution for a wide range of current and emerging artificial intelligence applications. The ARC EV7x Vision Processors integrate up to four enhanced vector processing units (VPUs) and a DNN accelerator with up to 14,080 MACs to deliver up to 35 TOPS performance in 16-nanometer (nm) FinFET process technologies under worst case conditions, 4X the performance of the ARC EV6x processors. In addition, the new EV7x design combines clock and power gating technologies with architectural enhancements to reduce power consumption. To speed application software development for ARC EV7x Vision Processors, Synopsys’ MetaWare EV Development Toolkit provides a comprehensive software programming environment based on common embedded vision standards, including OpenVX™ and OpenCL™ C. The combination of high-performance vision engine and DNN accelerator with high productivity programming tools make the ARC EV7x Embedded Vision Processors ideal for a broad range of vision applications including advanced driver assist systems (ADAS), video surveillance, smart home, and augmented and virtual reality.

New High-Performance Multicore Architecture with Deep Neural Network Accelerator

The DesignWare ARC EV7x Vision Processors’ heterogeneous multicore architecture includes up to four high-performance VPUs. Each EV7x VPU includes a 32-bit scalar unit and a 512-bit wide vector DSP and can be configured for 8-, 16-, or 32-bit operations to perform simultaneous multiply-accumulates on different streams of data. The optional DNN accelerator scales from 880 to 14,080 MACs and employs a specialized architecture for faster memory access, higher performance, and better power efficiency than alternative neural network IP. In addition to supporting convolutional neural networks (CNNs), the DNN accelerator supports batched LSTMs (long short-term memories) for applications that require time-based results, such as predicting the location of a pedestrian based on their observed path and speed. The vision engine and the DNN accelerator work on tasks in parallel, making the EV7x particularly efficient for autonomous vehicles and ADAS applications where vision algorithms operate concurrently.

Latest Advancements for Security and Safety

The DesignWare ARC EV7x Vision Processors offer optional AES-XTS encryption engines to protect data passing from on-chip memory to the vision engine and DNN accelerator. The AES-XTS engine prevents high-value data such as training datasets and personal biometric data, including facial recognition and retina scans, from being exploited. ASIL B or ASIL D Ready versions of the new processors, the ARC EV7xFS portfolio, accelerate ISO 26262 certification of automotive SoCs. The functional safety-enhanced processors offer hardware safety features, safety monitors, and lockstep capabilities that enable designers to achieve stringent levels of functional safety and fault coverage without significant impact on power or performance. In addition, a new Hybrid option enables system architects to select safety levels up to ASIL D in the software, post-silicon.

Enhanced High-Productivity Software Programming Environment

The ARC EV7x Vision Processor family is supported by MetaWare EV Development Toolkit, a comprehensive, high-productivity software development environment based on common embedded vision standards, including OpenVX™ and OpenCL™ C. The tool suite enables the development of efficient computer vision applications on the EV7x processor’s vision engine as well as automatic mapping and optimization of neural networks graphs on the dedicated DNN accelerator. The mapping tools support Caffe and Tensorflow frameworks, as well as the ONNX neural network interchange format.

DesignWare EV7x Processor Architecture

DesignWare ARC EV7x Vision Processor Block Diagram



DesignWare ARC EV7x Embedded Vision Processor IP
DesignWare ARC EV7xFS Vision Processor Datasheet

 

Highlights
Products
Downloads and Documentation
  • Optimized for high frame-rate and video resolution embedded vision and imaging applications
  • Integrates 32-bit scalar core, 512-bit vector DSP, and optional DNN accelerator
  • Configurations available with 1, 2 or 4 vector processing units and DNN performance options
  • DNN accelerator delivers up to 35 TOPS performance
  • DNN accelerator features advanced hardware and software techniques for improving performance and reducing bandwidth and latency
  • Works with all host processors to accelerate vision and deep learning algorithms
  • ASIL B or D Ready EV7xFS products integrate safety-critical hardware features while maintaining high performance and minimizing area and power
  • IEEE 754-compliant vector floating point unit option offers single precision or half precision operations and advanced math functions
  • Integrated AES encryption option protects valuable data such as training and sensor data from evolving threats without compromising performance
  • High productivity MetaWare EV Development Toolkit supports OpenCV, OpenVX and OpenCL C, Tensorflow, Caffe, and ONNX standards
EV71 processor for embedded vision applications with a single vector processing unitSTARs Subscribe
EV72 processor for embedded vision applications with two vector processing unitsSTARs Subscribe
EV74 processor for embedded vision applications with four vector processing unitsSTARs Subscribe
DNN 14080 MAC OptionSTARs Subscribe
1760 MAC DNN Option for EV Vision Processor IPSTARs Subscribe
3520 MAC DNN Option for EV Vision Processor IPSTARs Subscribe
DNN 7040 MAC OptionSTARs Subscribe
880 MAC DNN Option for EV Vision Processor IPSTARs Subscribe

Description: 1760 MAC DNN Option for EV Vision Processor IP
Name: dwc_ev_dnn_1760_mac_option
STARs: Open and/or Closed STARs
myDesignWare: Subscribe for Notifications
Product Type: DesignWare Cores
Documentation:
  
Description: 3520 MAC DNN Option for EV Vision Processor IP
Name: dwc_ev_dnn_3520_mac_option
STARs: Open and/or Closed STARs
myDesignWare: Subscribe for Notifications
Product Type: DesignWare Cores
Documentation:
  
Description: 880 MAC DNN Option for EV Vision Processor IP
Name: dwc_ev_dnn_880_mac_option
STARs: Open and/or Closed STARs
myDesignWare: Subscribe for Notifications
Product Type: DesignWare Cores
Documentation:
  
Description: DNN 14080 MAC Option
Name: dwc_ev_dnn_14K_mac_option
STARs: Open and/or Closed STARs
myDesignWare: Subscribe for Notifications
Product Type: DesignWare Cores
Documentation:
  
Description: DNN 7040 MAC Option
Name: dwc_ev_dnn_7k_mac_option
STARs: Open and/or Closed STARs
myDesignWare: Subscribe for Notifications
Product Type: DesignWare Cores
Documentation:
  
Description: EV71 processor for embedded vision applications with a single vector processing unit
Name: dwc_ev71_processor
STARs: Open and/or Closed STARs
myDesignWare: Subscribe for Notifications
Product Type: DesignWare Cores
Documentation:
  
Description: EV72 processor for embedded vision applications with two vector processing units
Name: dwc_ev72_processor
STARs: Open and/or Closed STARs
myDesignWare: Subscribe for Notifications
Product Type: DesignWare Cores
Documentation:
  
Description: EV74 processor for embedded vision applications with four vector processing units
Name: dwc_ev74_processor
STARs: Open and/or Closed STARs
myDesignWare: Subscribe for Notifications
Product Type: DesignWare Cores
Documentation: