Custom physical artificial intelligence chip startup SiMa.ai Technologies Inc. today announced it has raised $150 million in new funding at a $1.45 billion valuation to scale up its development environment and fund the next generation of its embedded AI compute platform.

Physical AI devices are hardware systems including robots, drones and autonomous vehicles that integrate artificial intelligence with sensors and actuators to reason about and take action in relation to the real world. The iconic example is a self-driving car that uses AI to “see” the world using cameras and laser imaging, stop at stoplights and work within traffic.

The global market for physical AI devices continues to surge as more connected machines are transformed into intelligent interfaces and SiMa.ai is building the hardware and software that will act as the “brains” for these machines.

“Physical AI in humanoids, automotive, and drones is the gateway to a $50 trillion market that has remained largely untouched by modern innovation,” said founder and Chief Executive Krishna Rangasayee. “While others are still figuring out the pieces or repurposing their cloud offerings, we’ve built the entire puzzle.”

Fidelity Management & Research Company and Amplify co-led the Series C financing round. Alter Venture Partners, Dell Technologies Capital and StepStone Group also participated in the round, alongside new investors AllianceBernstein, Baron Capital and J.P. Morgan.

Rangasayee said the funding would help “catapult” SiMa.ai into a dominant position by giving it the resources to scale its purpose-built platform and compete against incumbent rival Nvidia Corp., which produces more expensive CUDA-based hardware.

The capital raise will fund scaling Palette Neat, the company’s industry-first agentic development environment for Physical AI, and building the next generation of its system-on-chip, Modalix, intended for embedded robotics. SiMa.ai said it is aiming to deliver 1,000 tera operations per second with new hardware in the first half of 2028.

The highest-end Nvidia modules, the Jetson AGX Thor series, hit 2,000 FP4 TOPS and are designed for full-scale humanoid robots and complex autonomous machines. However, mainstream robotic and heavy drone modules such as Jetson T2000 and T3000 cover around 400 and 865 TOPS for mid-tier robotic arms and heavy delivery drones, maintaining a lower power tier to preserve battery life.

As the market-dominant physical AI software, Nvidia’s CUDA-based architecture is also among the most complex and power-hungry. SiMa.ai seeks to offer a lower-power alternative that includes custom-built silicon and software support that can deploy in days or hours.

The company said the next-gen hardware will let it aggressively expand into medium- to high-end drones, humanoid robotics, automotive advanced driver assistance systems and AI-powered cockpits.

Current customers and partners of SiMa.ai include ARK Electronics LLC, AverMedia Technologies Inc., Robert Bosch GmbH, Emerson Electric Co., Micron Technology Co. and Synopsys Inc.

Image: SiliconANGLE/Microsoft Designer