New Tensilica 5 DSP Enhances Automotive AI Audio Processing

NXP Semiconductors has made significant strides in the realm of digital signal processing (DSP) with the introduction of a cutting-edge chip designed specifically for AI-based audio applications in software defined vehicles (SDVs) and software defined radio (SDR).

The NXP SAF9xxx chip leverages the latest Tensilica 5 DSP core from Cadence, incorporating dedicated neural-network engines tailored for self-learning audio and speech functionalities. This innovative approach sets the SAF9xxx family apart, offering a versatile solution that caters to a wide range of audio and radio requirements.

One of the standout features of the SAF9xxx family is its software-defined radio option, which boasts up to five integrated tuners capable of covering various global broadcast radio standards such as DAB, HD-Radio, DRM, CDR, and AM/FM. This single chip device provides scalability and flexibility for audio and radio applications, making it a compelling choice for automotive and radio manufacturers.

With a focus on enhancing audio performance in automotive infotainment systems, the NXP SAF9xxx chip enables OEMs to differentiate their mid-range and high-end vehicles through advanced machine learning capabilities. By integrating neural networks, the chip can distinguish between drivers and passengers based on voice characteristics, while also detecting emergency vehicle sirens to provide crucial audiovisual alerts to drivers.

Moreover, the SAF9xxx family is optimized for low-latency audio applications, enabling real-time active noise cancellation for engine, road, wind, and tire noise. Its Gigahertz-DSP performance, coupled with a high-speed onboard memory and a comprehensive set of analog and digital I/Os, ensures seamless support for demanding audio features, enhancing the overall in-car audio experience.

John van den Braak, Senior Vice President and General Manager for Audio Systems at NXP, emphasized the chip's readiness as an infotainment solution, highlighting its AI/ML processing capabilities that deliver enhanced functionality, performance, and features for current and future software-defined vehicles. The chip's scalable architecture further simplifies integration for OEMs, making it a cost-effective choice for vehicle design.