AI chips, which are semiconductors designed to accelerate machine learning, have many applications. One of the promising use cases, according to Albert Liu, is using AI chips in autonomous driving vehicles.
That’s why Liu’s AI chipmaking startup Kneron has been quietly racking up investments to march into smart transportation. It recently closed a new round of $25 million funding led by Lite-On Technology, a Taiwanese optoelectronic pioneer, which was a strategic investor in the round. Other investors included Alltek, PalPilot, Sand Hill Angels and Gaingels.
The new proceeds lifted Kneron’s total funding to over $125 million since its inception in 2015. The San Diego- and Taipei-based company has assembled a list of renowned investors, including Hong Kong tycoon Li Ka-Shing’s Horizon Ventures, Alibaba, Qualcomm, Sequoia and Foxconn, the world’s largest electronics manufacturer and Apple’s supplier.
Liu previously told TechCrunch that the company was expected to turn profitable in 2023, which would be a “good time” to go public. When asked more recently, the founder was more tight-lipped about the firm’s IPO prospects, though he said the listing would likely be in the U.S.
“The future of L4 or L5 autonomous driving lies not only in the vehicle but also roadside AI,” Liu, who worked on computer vision at Qualcomm, told TechCrunch in an interview.
Kneron’s latest investment followed the release of its first auto-grade semiconductors for advanced driver-assistance systems (ADAS) and autonomous vehicles.
With roadside perception units that can communicate with nearby vehicles, an ambulance would not have to stop at intersections, Liu explained. Such infrastructure is particularly useful in Asian countries where traffic conditions are more complicated than those in the U.S., he added.
Through the strategic investment, Kneron and Lite-On will jointly develop roadside AI boxes that will be powered by Kneron’s edge AI chips.
The startup’s chips are “reconfigurable”, meaning they combine the flexibility of software and the high speed of the hardware. In the case of AV, its silicon can be used for both the large AI engine inside a vehicle and powering the small sensors layered on the exterior of the car, explained Liu and as shown in this demo video:
Kneron is currently pulling in $3-4 million in monthly revenue from its pool of 30 enterprise customers; 30-40% of its revenues come from the U.S. at the moment.
The company has been forming deep alliances with industry partners. In May, Kneron agreed to buy image signal processor Vatics from Vivotek, a subsidiary of Delta Electronics. Kneron counts Foxconn as a strategic investor, which has been using the startup’s chips in its “MIH” manufacturing platform for electric vehicles.