Scalable Vector Signal Processor IP Takes On HD Video—And High-End Audio

April 14, 2005
Processing high-quality video and audio data streams in real time requires plenty of compute horsepower. To meet that throughput demand, OnDemand's Vector Signal Processor (VSP) core offers a scalable, synthesizable compute engine that handles H.264

Processing high-quality video and audio data streams in real time requires plenty of compute horsepower. To meet that throughput demand, OnDemand's Vector Signal Processor (VSP) core offers a scalable, synthesizable compute engine that handles H.264 video compression up through the extended HDTV profile.

The architecture can execute functions such as low-density parity-check (LDPC) coding, which is a subclass of forward-error-correction coding. Such algorithms help minimize errors in high-speed data-transmission systems. The VSP itself consists of a parallel processing engine based on multiple identical compute slices.

These slices can operate in either a single-instruction/multiple-data or multiple-instruction/multiple-data (SIMD or MIMD) mode (see the figure). Each slice contains two data memories, an address generator, a register file, and a calculation unit. This unit includes a dual multiplier-accumulator that can handle 8-, 16-, 24-, or 32-bit operations, as well as shifters, logical entities, alignment, and rounding units.

Furthermore, these slices are interconnected through a global data bus and switch matrix that eliminates bus congestion and maximizes data movement. A global arithmetic unit enables synchronous processing of partial results from several slices. The global data bus links all slices, the global arithmetic unit, and the I/O ports.

The scalable core's program sequencer houses the program counter, which addresses the very long instruction words (VLIWs) held in the program memory. Each slice is controlled by its own subword in the VLIW instruction.

Global instructions (part of the VLIW instructions) manage global data transfers, jumps, and loops. Because each slice has its own resources and can perform a different instruction sequence, it's possible to execute different algorithms in parallel.

The VSP integrated development environment lets designers easily configure a VSP-based solution and develop the VSP software. In turn, this would support the creation of a core with the necessary number of slices to handle the desired application.

The tool suite includes VSP-Pilot, a development and simulation environment. Its VSP-Target chip integration tool, which integrates the VSP core within CAD environments, automatically generates testbenches. Other suite features include VSP-Board, a real-time evaluation board, and VSP-Courier, a configuration tool. A comprehensive C-based library provides proven algorithm implementations (FFT, DCT, Viterbi, etc.) to speed signal-processing software creation.

Contact OnDemand for licensing terms.

OnDemand Microelectronicswww.ondemand.co.at
About the Author

Dave Bursky | Technologist

Dave Bursky, the founder of New Ideas in Communications, a publication website featuring the blog column Chipnastics – the Art and Science of Chip Design. He is also president of PRN Engineering, a technical writing and market consulting company. Prior to these organizations, he spent about a dozen years as a contributing editor to Chip Design magazine. Concurrent with Chip Design, he was also the technical editorial manager at Maxim Integrated Products, and prior to Maxim, Dave spent over 35 years working as an engineer for the U.S. Army Electronics Command and an editor with Electronic Design Magazine.

Sponsored Recommendations

Comments

To join the conversation, and become an exclusive member of Electronic Design, create an account today!