Syntiant and Infineon Offer Always-on Ultra-Low-Power Local Voice Control

December 5 2019, 10:00
Syntiant, a California-based AI chip company focused on moving machine learning from the cloud to edge devices, announced a partnership with Infineon Technologies, to service the rising demand for voice user interface (VUI) in electronic devices. Optimized for battery-powered devices from true wireless earbuds to mobile phones to smart speakers, Syntiant’s partnership with Infineon enables more efficient voice control and custom command detection to numerous product configurations.

As a member of Infineon’s VUI partner network, Syntiant’s microwatt-power Neural Decision Processors (NDPs) combined with Infineon’s XENSIV MEMS microphones are able to deliver crystal clear voice commands to edge applications with ultra-low power consumption, free from cloud connectivity. The joint solution enables developers to easily create custom, always-on voice user interfaces, allowing exact voice pick-up from long-distance, with whispered speech or in noisy environments. 
 
"Infineon´s best-in-class XENSIV MEMS microphones are redefining the acoustic, electric power consumption and robustness standards for voice user interfaces," says Dr. Roland Helm, Ph.D., head of product line sensors at Infineon Technologies. "Together with Syntiant we are facilitating clear voice commands at the edge that support audio and keyword classifications, on and off-line, in a variety of challenging, power-critical always-on and battery driven environments."
 

The Syntiant NDP9101B0 development platform combines Infineon's high performance digital XENSIV IM69D120 MEMS microphone with the Syntiant NDP101 Neural Decision Processor (pictured above) to prototype applications such as keyword spotting, wake word processing and speaker identification that consume only 140µW. Other development capabilities include audio event and environment classification, and sensor analytics.
 
"By pairing Infineon’s advanced microphone technology with our NDPs, we can deliver low latency, always-on voice applications that are extremely low power and highly responsive to varied noise and audio inputs," says Kurt Busch, CEO of Syntiant. "Developers can use our NDP9101B0 Pi platform to quickly create their own custom voice-based user interface for local commands without cloud connection that will protect the privacy and security of end users."

Providing 20x the throughput of the typical low-power MCU solution, the Syntiant NDP can support up to 63 local voice commands. Syntiant’s processor also comes with the Amazon Alexa Voice Service keyword model which enables a close-talk built-in Alexa experience. Infineon’s XENSIV IM69D120 MEMS microphone is designed for applications where low self-noise, wide dynamic range, low distortions and a high acoustic overload point are required. 
 

The NDP9101B0 platform is implemented as a Raspberry Pi shield with easily configurable jumpers to connect a number of different microphones and sensors. Other key features include 500KB of serial flash, dual on board IM69D120 Infineon high dynamic range microphones, pre-configured micro SD card, on board programmable Si5351A clock IC and four LEDs for IO display.

Syntiant has been named a CES 2020 Best of Innovation Awards Honoree for its Syntiant NDP100 microwatt-power Neural Decision Processor. The NDP100 provides always-on AI processing for voice and other sensor applications in battery-powered devices. Custom built to run neural workloads, the NDP100 processor consumes less than 150 microwatts, enabling devices to be approximately 200x more efficient, and provides 20x the throughput of the typical low-power MCU solution.
www.syntiant.com | www.infineon.com
related items