Audio DSPs bring AI audio to automotive infotainment

NXP’s SAF9xxx audio DSPs, with integrated NNEs and hardware-based accelerators, bring AI/ML audio capabilities to automotive infotainment.

NXP Semiconductors has launched its SAF9xxx family of audio DSPs , bringing AI audio capabilities to automotive infotainment. Called a significant improvement in automotive audio processing technology, the new solution supports AI-based audio in software-defined vehicles (SDVs), delivering a differentiator for automotive infotainment in mid-range and high-end vehicles.

The SAF9xxx family, using Cadence’s high-performance Tensilica HiFi 5 DSPs, combined with dedicated neural-network engines (NNEs), enables the implementation of next-generation self-learning audio and speech applications. It also provides a software-defined radio option with up to five integrated tuners, covering all major global broadcast radio use cases, including DAB, HD-Radio, DRM, CDR, and AM/FM, providing a scalable architecture with one chip.

NXP said the SAF9xxx’s machine learning (ML) capabilities with neural network integration deliver new functionalities. One application cited is using the solution to recognize and distinguish the driver from passengers by identifying their voice pitches and accents. Also, emergency vehicle sirens can be detected with MLs algorithms and used to provide drivers with audiovisual information about the location and movement of an emergency vehicle.

The SAF9xxx family, housed in a 9 × 9-mm FC-CSP package, is optimized for low-latency audio applications, enabling real-time active cancellations of engine, road, wind and tire noise.?Offering Gigahertz-DSP performance, these devices also feature large, fast-access onboard memory; a comprehensive set of analog and digital I/Os; and PCIe and Gigabit-Ethernet connectivity.

NXP will unveil the SAF9xxx family at the Audio Engineering Society 5th?International Conference on Automotive Audio , June 26-28, 2024.?NXP also announced a large audio IP partner ecosystem with a range of audio software and algorithms.