SiMa.ai, a developer of chips and software for running AI directly on physical devices, has raised a $150 million Series C at a $1.45 billion valuation, TechCrunch reported Monday. Fidelity Management & Research Company and Amplify co-led the round, lifting the company's total funding above $500 million.

What Happened

Alter Venture Partners, Dell Technologies Capital and StepStone Group also took part in the round. The new valuation is a step up from SiMa.ai's Series B in July 2025, when it raised $85 million at a $960 million valuation, according to PitchBook data cited by TechCrunch.

The company was founded in 2018 and is led by founder Krishna Rangasayee, who was previously chief operating officer at chipmaker Groq. SiMa.ai designs energy-efficient chips, along with the software to program them, that let robots, drones, cameras and other machines run AI models locally rather than sending data to cloud servers for processing.

RelatedWaymo Lines Up $5 Billion Loan From Blackstone, PIMCO and Other Lenders in First Debt Raise

Its pitch rests on two advantages over Nvidia's GPUs for this class of work: lower latency, because inference happens on the device, and lower cost. TechCrunch said the company is targeting the physical AI market, including humanoid robots.

Why It Matters

Physical AI, in which models perceive and act in the real world through machines, places different demands on hardware than data-center training. A robot or drone has a limited power budget, often operates with unreliable connectivity, and needs to react in milliseconds. Sending camera feeds to the cloud adds delay and bandwidth cost, and can raise privacy concerns when the footage involves people or sensitive sites. That is the gap SiMa.ai and other edge-inference chipmakers are trying to fill with purpose-built silicon.

The round shows that investors are willing to back specialist chip designers in that niche despite Nvidia's dominance in AI hardware. The investor list mixes large asset managers, in Fidelity and StepStone, with Dell Technologies Capital, the venture arm of a company that sells servers and edge computing gear to enterprises.

Context

The funding lands during a wave of investment in physical AI. On the same day, AMD agreed to buy Fei-Fei Li's World Labs, which builds world models used to simulate environments for robot training, for $8.2 billion. Much of the money in the sector is flowing to software and models that train robots, but running those models on deployed machines still requires inference hardware that fits within tight power and cost limits.

SiMa.ai's valuation climbed by about 50% between rounds, over roughly 14 months. By comparison, consumer AI agent startup Instinct, which also announced funding Monday, quadrupled its valuation to $10 billion in about a month.

What to Watch

TechCrunch's report did not set out specific product or customer milestones tied to the new money. Signals to track include which robot and drone makers adopt SiMa.ai's chips in shipping products, how its performance compares with Nvidia's edge platforms in independent tests, and whether the company raises again as physical AI spending grows.