Recogni Inc., a startup developing artificial intelligence hardware for autonomous vehicles, today announced that it has closed a $48.9 million funding round from a group of investment firms and auto industry players.

The announcement is the latest sign that investors’ interest in the autonomous vehicle market continues to be strong. 

The round was led by WRVI Capital, which counts Nuvia Inc. among its other portfolio companies. Nuvia is a central processing unit startup that last month agreed to be acquired by Qualcomm Inc. for $1.4 billion. WRVI was joined in Recogni’s funding round by Mayfield Fund, Toyota Motor Corp., BMW AG and others.

San Jose, California-based Recogni is developing an AI inference module that can be installed inside a vehicle to run its autonomous driving software. The module is based on a custom application-specific integrated circuit designed by the startup’s engineers.

Recogni claimed when it launched from stealth in 2019 that its hardware was up to 500 times more power-efficient than competing systems. In a post on the startup’s blog dated November 2020, product manager Sidhart Krishnamurthi elaborated that the Recogni module is capable of providing 100 teraflops of performance per watt. A single teraflop equals a trillion computations per second.

Recogni says that its AI module’s energy efficiency can simplify the design of electric cars with autonomous driving features. Driving without human input requires a vehicle to process data from multiple sensors in near-real time, which consumes a significant amount of power.

That, in turn, reduces the vehicle’s battery life. Technologies such as Recogni’s module that promise to run AI software more efficiently could enable automakers to build autonomous vehicles capable of covering more distance per charge.

Another benefit Recogni says its module enables: the ability to replace lidar sensors with stereo cameras. Stereo cameras, which can collect depth information, are theoretically more cost-efficient than the lidar sensors used in most autonomous vehicles today. But they’re not used often because they require significant computing resources to support. Recogni says its module’s ability to provide 100 teraflops of performance per watt turns stereo cameras into a more viable option for automakers.

The startup will use the new $48.9 million funding round to bring its AI module to market, as well as expand its sales and engineering teams. The round follows a $25 million raise in July 2019 that also saw the participation of Toyota and BMW.

“We are moving along our journey of solving the challenge of perception processing and power efficiency, by building the world’s highest-performing AI inference system at the lowest energy consumption,” said Recogni Chief Executive Officer R K Anand.

Firms developing technology for autonomous vehicles are drawing significant interest from investors. Several lidar sensor makers, among them Innoviz Technologies Ltd and Ouster Inc., have announced plans in recent months to go public through mergers with special-purpose acquisition companies. More recently, General Motors Co.’s Cruise autonomous vehicle unit received a more than $2 billion investment from Microsoft Corp. and others. 

Photo: Unsplash