AI chip startup SiMa.ai launches auto business with former Bosch, Mercedes executive

SiMa.ai CEO Krishna Rangasayee poses for a picture with his machine learning system used for processing video and images for security cameras, satellite images, drones, industrial robots, and eventually self-driving cars in, San Jose, California, U.S., in this undated handout photo. SiMa.ai/Handout via REUTERS/File Photo

OAKLAND, Calif., Nov 3 (Reuters) - Silicon Valley-based AI chip startup SiMa.ai on Thursday said it is entering the automotive industry and has recruited Harald Kroeger, a former executive at top tier auto supplier Bosch and Mercedes Benz (MBGn.DE), to lead that business.

Kroeger, who is on the board of electric pickup truck firm Rivian (RIVN.O), will join SiMa.ai’s board, which includes top chip industry executives Moshe Gavrielov, a Taiwan Semiconductor Manufacturing Co. (2330.TW) board member, and Lip‑Bu Tan, an Intel (INTC.O) board member.

The announcement comes as chip makers' competition in the automotive industry intensifies with Nvidia Corp (NVDA.O), Qualcomm Inc (QCOM.O), and Intel’s recently listed Mobileye (MBLY.O) all vying for car makers.

SiMa.ai’s main product called MLSoC, short for machine learning system on chip, processes video and images, an important technology for assisted and self-driving technology.

"I was on the knife's edge," said Krishna Rangasayee, CEO of SiMa.ai about the decision to enter the automotive market which he noted can take years to generate revenue from.

“Automotive is 40% of the overall market from a computer vision perspective...Now that we have established our technology, I thought it was the right bet to focus on automotive.”

Kroeger said the focus would be on assisted driving rather than self-driving technology. But SiMa.ai has also been talking to some self-driving tech firms, said Rangasayee.

Even in assisted driving, as car makers want to offer technology that does more than just keep the lanes or distance between cars, Kroeger said a more advanced machine learning-based technology was needed and a more power efficient chip was also key.

"If you looked at today's products in the market right now, some of those things in the cars use up to two kilowatts of power, which means you could fry some eggs in the car on that dissipation power," he said.

Reporting By Jane Lanhee Lee; Editing by Kirsten Donovan

Our Standards: The Thomson Reuters Trust Principles.