Knowles AISonic IA8201 Raspberry Pi development kit is designed to bring voice, audio edge processing, and machine learning (ML) listening capabilities to various systems, and can be used to evaluate the company’s AISonic IA8201 DSP that was introduced about two years ago.
The kit is comprised of three boards with an adapter board with three buttons connecting to the Raspberry Pi, as well as the AISonic IA8210 DSP board itself connected via a flat cable to a microphone array.
Knowles did not provide the full details for the development but says it enables wake-on-voice processing for low latency voice UI, noise reduction, context awareness, and accelerated machine learning inferencing for edge processing of sensor inputs.
Some of the use cases include Low Power Voice Wake to listen for specific OEM keywords to wake the host processor, Proximity Detection when combined with an ultrasonic capable speaker and microphone, Security System for example to detect glass breakage, smoke alarms, etc…
Programming is performed through the “DSP SDK” with Knowles and Xtensa HiFi 3 instruction sets enabling voice and audio processing, voice user interface, and ambient sound processing. There’s also a partner program called “Open DSP” bringing third-party algorithms to the ecosystem. The myKnowles portal is supposed to offer resources, including solution kits, product briefs, discussion forums, and more. But it requires registration, which I did, and manual approval, which I’m still waiting for.
The press release does offer a few more insights with regards to software, notably support for Amazon Alexa, Sensory, Retune, and Alango voice services, as well as TensorFlow Lite Micro SDK for AI/ML applications. We also learn the kit actually comes with two microphone arrays, one with two microphones, and the other with three microphones in order to let developers evaluate and select the optimal solution for their specific application.
While we have limited information about the development kit itself, there are mode details about the AISonic IA8201 DSP with the following specifications:
- MCU Cores
- “DeltaMax” Xtensa LX5 128-bit 4-way floating-point SIMD DSP core, with Knowles instruction set extensions optimized for frame-based audio processing and DNN acceleration
- “HemiDelta” Xtensa LX5 64-bit 2-way floating-point SMID DSP core, with both Xtensa HiFi 3 and Knowles instruction sets, optimized for low power
- Memory – 1.44MB RAM including 1MB available to the user
- Audio Interfaces
- Up to 4x PDM Digital Microphones – 1 stereo input, 4x mono inputs, and 1 stereo output, supporting clock rates up to 6.144 MHz
- Up to 3x I 2 S/TDM ports supporting 8 channels each of 32-bit audio data using a 24.576 MHz input clock
- Control Interfaces – SPI, I 2 C, UART, available GPIOs.
- Clock up to 175 MHz
- Voltage supply requirements
- IA8201BC – 1.8V Vdd
- IA8201CQ – 1.8V Vdd and 3.3V Vdd
- Temperature Range -20 to 85°C
- Packaging Options
- eWLB 3.00×2.6×0.715mm, 0.4 pitch, 42 ball (IA8201BC)
- QFN 6.00×6.00×0.75mm, 0.5 pitch, 40 lead (IA8201CQ)
The DSP and devkit can be used for a variety of applications including the Smart Home, home entertainment, Smart Buildings, residential and commercial security, as well as for inference engines for the industrial and commercial sectors.
The Knowles IA8201 Raspberry Pi Development Kit is now available for order, but you’d have to contact the company if you are interested, AFAICS it’s not sold online. More details can be found on the devkit and AiSonic processors pages, with the latter, also including AISonic IA8508 quad-core DSP adding Cortex-M4 and SSP (single sample processor for echo cancellation, sample rate conversion) cores to the DeltaMax and HemiMax cores of IA8201 DSP.
Jean-Luc started CNX Software in 2010 as a part-time endeavor, before quitting his job as a software engineering manager, and starting to write daily news, and reviews full time later in 2011.