Processor IP

Originally developed in 1997 and perfected over 4 generations, our processor architecture is the most power-efficient and highest-performing video/vision DSP that you can license on the market. Compared to CPUs or GPUs, the architecture’s power consumption is 1000x lower, while improving performance per silicon dollar more than 100x. The multicore v-MP4000HDX architecture scales from ultra-low-cost single-core applications to high-performance, many-core applications, and supports both video coding as well as computer vision acceleration.

Unified video/vision architecture enables flexibility

Videantis architecture

The videantis v-MP4000HDX architecture runs a wide variety of video and image processing tasks very efficiently. Whether it’s video compression or decompression, computer vision, or imaging algorithms, the videantis architecture provides a very efficient implementation. The heterogeneous multicore architecture combines multiple VLIW/SIMD media processors (v-MPs) with a number of stream processors (v-SPs) that accelerate bitstream packing and unpacking. Each processor includes its own multi-channel DMA engine for efficient data movement.

Low power extends battery life and prevents overheating

ice-463845_640

Low energy consumption extends battery life of mobile devices. Low power consumption also prevents camera modules and image sensors from overheating in automotive and surveillance applications. A new class of computer vision applications, such as gesture interfaces or face analysis, require always-on operation, again necessitating ultra low power consumption. The v-MP4000HDX processor has been architected for low-power operation from the ground up, resulting in a 1000x power reduction compared to other solutions on the market.

High-performance enables new applications

gear-192875_640

Computer vision applications, 4K or Full HD video coding, and high-resolution image processing require tremendous amounts of processing. With performance of well over 1000 GOPS, the v-MP4000HDX processor provides ample computational horsepower. The heterogeneous, multicore VLIW/SIMD architecture was designed specifically for vision algorithms and complex video applications, resulting in a 100x performance improvement compared to other processors.

Single scalable architecture

Scalable architectureWe introduced our first multicore offering to the market at the Microprocessor Forum in 2005, and have over 10 years of experience in optimizing and parallelizing software for efficient multicore operation. The single scalable architecture enables different operating points across your product lines: from low-cost to high-performance. Our software is fully scalable too, automatically using all processor cores that are available for the task. Our embedded vision and video coding libraries are implemented in such a way that they take full advantage of the number of videantis cores available in the system.

Software programmable

Our processor architecture is fully software programmable. Firmware can be upgraded to cover additional functionality. You can keep your SOCs in the market longer, or you can expand your target market and address a wide variety of customer requests. Firmware upgradeability also reduces risk. In case interoperability issues between your SOC and other chips in the system are found, these can be solved by adjusting the software, without costly silicon respins. All of our customers have always come back to us with requests for additional software, expanding their time-in-market, and capturing more customers.

Learn more?

Read about our computer vision and video coding optimized libraries, or feel free contact us for more information.