DSP Concepts Announces Multicore Support for Audio Weaver DSP Programming Software

December 15 2016, 03:00
DSP Concepts, a developer of embedded audio DSP solutions, has announced multicore support for its Audio Weaver modular programming software. Audio Weaver can now operate in heterogeneous SoC (System on a Chip) devices that employ multiple cores of different types. The new capability extends Audio Weaver’s user-friendly programming environment into more applications, including complex signal processing chains in automotive environments.

Audio Weaver is a popular product-development system for audio-enabled products. The system is built around two components, the Audio Weaver Runtime Core, a dynamically-configurable audio-processing engine, easy to integrate into production software; and Audio Weaver Designer, a PC graphical configuration and tuning tool. It enables algorithm and product developers to more quickly and efficiently develop products and technology.

The Audio Weaver programming environment is a modular, GUI-based environment that allows any audio engineer to build and optimize complex signal processing chains without the need for DSP programming skills, and it can now be used in several leading SoCs, including products from Analog Devices (SC58x & SC57x SHARC), Renesas Electronics (R-car3 HiFi2) and Texas Instruments (Jacinto 6 C66x).

SoCs are finding increasing use in many complex systems — particularly in automotive applications where a system may have to accommodate microphone, camera and sensor input, voice command, active road noise cancelling, engine sound synthesis and other tasks in addition to traditional audio and video entertainment. Audio Weaver now supports a mix of processor cores including traditional DSPs, application processors, microcontrollers and even low power DSPs — all configured from an easy-to-use graphical interface.

“With multicore support, we are able to distribute the processing load among multiple cores and take full advantage of the latest SoCs’ capabilities,” says Dr. Paul Beckmann, Founder and CEO of DSP Concepts, Inc. “Heterogeneous support is critical because many SoCs contain a variety of processor types. Typical SoC applications require engineers to distribute audio processing tasks based on the unique features of each core, matching the right processor for the right purpose to ensure the most efficient processing and full use of the SOC’s capabilities.”

SoCs with multiple processing cores are increasingly common, as semiconductor manufacturers find it easier to increase performance by adding additional cores instead of increasing clock speeds. Yet the more cores an SoC incorporates, and the more variation in core configurations used by different manufacturers, the greater the programming challenge for OEMs and Tier 1 suppliers. As the number of cores increases, it becomes more difficult to assess performance as workloads are shifted among different cores. Audio Weaver’s real-time profiling feature allows instant evaluation of different programming options, so performance can be optimized and heat buildup minimized.

According to market analyst Richard Robinson, Director, Automotive Multimedia and Communications Service at Strategy Analytics, demand for SoCs in the automotive market is expected to grow as the use of networking and Internet connectivity in cars creates new possibilities and increased consumer demand for in-car technology. “The demands on in-car audio systems are expected to increase due to a number of features coming into the vehicle,” Robinson said. “These include more playback channels for 3D sound, which would include ADAS [Advanced Driver Assistance Systems] audio warnings; additional microphone processing for voice command, telephony, and active noise cancellation features; and more use of acoustical sensors for car systems failure diagnostics, EMS detection (type, distance and direction), and emergencies such as collisions and break-ins.”

“TI is excited to be one of the first infotainment SoCs to have multicore support from DSP Concepts’ Audio Weaver tool,” says Brad Ballard, Business Development Manager, Texas Instruments. “Our Jacinto 6 (DRA74x) family of infotainment processors is based on a heterogeneous, multicore architecture utilizing TI’s C66x DSPs for signal processing, and ARM Cortex-A15 CPUs for core applications and signal processing, and Cortex M4s for network control and other real-time tasks. DSP Concepts’ Audio Weaver tool with multicore support greatly simplifies the task of utilizing the potential of these types of multicore solutions, reducing development time and cost.”

“Audio signal processing continues advancing to find new solutions to difficult problems and to bring us better experiences in amazing ways. This is a trend that will continue,” says Colin Duggan, Director of Applications and Marketing Processor Technology Group, Analog Devices. “However, with this trend comes implementation complexity and requirements for higher performance. Audio Weaver with multicore support addresses both challenges by reducing the complexity of implementing audio signal processing algorithms and by supporting higher levels of performance through utilization of multiple cores. Audio Weaver paired with Analog Devices SHARC + ARM processors, such as the SC58x and SC57x portfolios, is a great combination for the latest and newly emerging audio signal processing advancements.”

“Audio Weaver is a feature that has deep roots in automotive from a playback standpoint, but we are just beginning to see the new microphone-based feature emerge, which will require more processing capabilities,” adds Amrit Vivekanand, Vice President, Automotive Business Unit, Renesas Electronics America, Inc. “DSP Concepts’ tool will enable automotive designers to unlock the full potential of our R-Car H3 SoC’s rich CPU and DSP engine offerings and allow the superior levels of integration for advanced audio features in-cabin such as surround sound, ADAS audio warnings, noise cancellation, and telephony.”
related items