Ideal for camera image processing, video post processing, gesture recognition, automotive driver assistance and computer vision applications, the IVP-EP core is based on a new, optimized architecture and is available both as a configurable core as well as a complete pre-built subsystem for easier integration into SoCs.

Compared to its predecessor, the IVP-EP processor gives customers up to 4X better performance for applications such as face detection, gesture recognition, augmented reality, video stabilization, high dynamic range (HDR) photos, HDR video, tracking, digital zoom, automotive lane departure, object detection and other imaging and video applications.

This enables sustained, complex video processing that cannot be achieved by other general-purpose CPU or GPU processors, with sufficiently low power dissipation for use in smartphones, tablets and other mobile devices. The IVP-EP core is scheduled to be available in May 2014.

“Mobile imaging algorithms demand extreme performance and power efficiency, as there is a high demand for better photos or videos as well as real-time image and video processing,” said Jack Guedj, Corporate VP, Intellectual Property Group, Cadence.

“We are moving quickly to offer increasingly flexible and higher performance imaging dataplane processor IP to keep up with the high-performance vision applications demanded by leading camera systems while also striving to deliver these solutions at much lower power than what was possible before.” said Mr. Guedj.

Features

-Based on a 4-way VLIW (very long instruction word) architecture

-With a 32-way vector SIMD (single instruction, multiple data) dataset

-Integrated DMA (direct memory access) transfer engine with up to 10GBps

-Imaging-specific operations to accelerate 8-, 16- and 32-bit pixel data types

-4X better performance than previous IVP core