MVE™ System-on-Chip

The Multi-core Video Analytics Engine (MVE™) is an easily configurable, compact, high-performance processing architecture that can be used to implement complete video analytics solutions in a single FPGA embedded in intelligent surveillance cameras.

MVE™ System-on-Chip (SoC) combines an inherently parallel multi-core processing architecture, and embedded complex video analytics algorithms in a configurable building block, simplifying hardware design and programming effort resulting in reduced development cost and time to market.

Embedded solutions also benefit from low-cost and low-power implementation when embedding the entire MVE™ into a single FPGA. This initial commercial implementation of MVE™ is on the Xilinx Spartan-3A DSP 3400A™ FPGA chip.

MVE™ SoC Implementation Statistics:

MVE™ is currently implemented in a Xilinx Spartan-3A DSP 3400A™ FPGA within a System-on-Chip (SoC) reference design with all required I/O functions for communication and data streaming (see block diagrams). The entire SoC solution uses 98% of the logic slices, 92% of the block RAMs and 42% of the DSP slices. The MVE™ analytics engine itself (excluding the MPMC-PLB part of the backbone and specialized I/O components) utilizes only 46% of the logic slices, 31% of the block RAMs and 30% of the DSP slices.

Key MVE™ Features:

• Stand-alone configurable Video Analytics engine based on high-performance C-MVA™ processor architecture
  
• Fully configurable VA functions such as people counting, detection of abandoned/stolen objects, entering forbidden zone, loitering, movement in wrong direction
  
• Enables single-chip VA solution in Xilinx FPGAs Image capture and video analysis done in HD image resolution at 30 frames per second
  
• Arbitrary number of events detected in parallel
  
• Pixel processing and acceleration implemented by parallel IP cores
  
• Reference design and evaluation kit available for the Eutecus Bi-i™ V401X VA platform and the Xilinx XtremeDSP™ Video Starter Kit (VSK)