Recogni Inc., a startup developing artificial intelligence hardware for autonomous vehicles, today announced that it has raised $102 million in a funding round co-led by Celesta Capital and GreatPoint Ventures.

The Series C round was also joined by existing and new investors including Mayfield, DNS Capital, BMW i Ventures Inc. and the Public Investment Fund’s Tasaru Mobility Investments.

Founded in 2017, Recogni develops AI hardware that allows autonomous vehicles to understand surrounding environments visually through raw data from sensors. Object detection requires a great deal of data taken from video and other sensors and Recogni’s hardware allows for more rapid recognition. The company said its chips allow for higher computational power at lower energy use, meaning higher efficiency.

Many AI chips on the market, such as graphical processing units, supplied by other major AI hardware manufacturers, such as Nvidia Corp., focus on consuming massive amounts of raw data in order to assist with developing and training AI models. However, Recogni’s chips work on the other end of the AI development spectrum called interference, where the AI model is deployed and is used to solve tasks or make predictions based on live data.

For generative AI models, inference occurs when a chatbot answers a question or summarizes an article for a user, when a model identifies an image of a dog as a dog or when an AI model classifies spam email. For example, when driving a vehicle would need to make inferences based on its surroundings, such as the road markings, signs and other vehicles visible to its sensors on the road in order to adjust its behavior.

As a result of its efficient low-power solution, the company said Recogni’s hardware solution is viable for more than just self-driving vehicles but can be deployed across numerous industries such as biopharmaceutical, financial, healthcare, retail, entertainment and fraud detection.

“The critical need for solutions that directly address the key challenges in AI inference processing — compute capability, scalability, accuracy and energy savings — is more urgent than ever,” Chief Executive Marc Bolitho. “Recogni is leading this transformative wave, engineering pivotal advances that will redefine data centers and enterprise and revolutionize industries like automotive and aerospace.”

Recogni launched its hardware solution, the Scorpio, its first low-power compute chip for autonomous vehicles, in December 2022. It is capable of producing 1,000 teraflops of inference for autonomous mobility, the company said, at the lowest power in the industry — with 10 to 20 times more power efficiency than competing solutions at the time.

“Generative AI systems in the market today are highly inefficient and consume too much power while adding system complexity, but Recogni’s proven technology is raising the bar on power performance to address the large compute needs of AI workloads,” said Sriram Viswanathan, Celesta Capital founding managing partner and chairman of Recogni’s board of directors.

Image: Pixabay