Ultra–low power AI accelerator startup Syntiant has closed a funding round of $55 million, bringing the company’s total raise to around $100 million. The company, which was founded in 2017, was backed by new and existing investors.
Syntiant also announced that it has shipped more than 20 million of its neural decision processor chips to date, making it one of the few edge AI chip startups shipping in substantial volumes today.
Among five new investors this round was Renesas Electronics Corp. Syntiant and Renesas have been collaborating on a “voice–controlled multimodal AI solution” since last summer, which combines Renesas’ RZ/V series microprocessors with Syntiant’s NDP120. This design uses Syntiant’s always–on neural decision processor to enable voice activation for a variety of vision–based AI applications, which are performed on the Renesas chip; all while keeping the standby power very low.
“By leveraging Syntiant’s leadership position in deep learning edge compute, our collaborations are enhancing user experiences through advanced AI–based voice and vision processing solutions across many use cases, such as security systems, personal devices, industrial and manufacturing, transportation, and logistics,” said Sailesh Chittipeddi, executive vice president and general manager of Renesas’ IoT and infrastructure business unit, in a statement.
The NDP120 is an SoC featuring Syntiant’s second generation AI accelerator core alongside a Tensilica HiFi3 DSP for feature extraction and an Arm Cortex–M0 core for system management. The second–generation Syntiant core offers scope for bigger neural networks than its predecessor, including running multiple neural networks simultaneously. This might include keyword/wake word detection, speaker ID and/or command detection.
“The NDP120 can bring the level of performance that you would typically find in a plugged–in smart speaker to a battery powered device, that’s really the goal for this product,” Syntiant CEO Kurt Busch told EE Times in a previous interview. “The NDP120 is primarily for audio [processing] as well as sensor fusion, with one or more of those sensors being a microphone… for far–field and noise clean up. It’s ideal for [AI–powered] noise reduction.”
Syntiant said the recent round of funding will be used to accelerate deployment of its third–generation neural decision processor core, as well as continuing work on software development. The company expects to debut third–gen products next year.