SiMa.ai Technologies Inc., a developer of custom physical AI chips, has closed a $150 million funding round to accelerate development of its next-generation embedded AI platform. The investment aims to expand their agentic development environment and deliver powerful, low-power silicon for a growing global market of intelligent hardware systems.
- Raised $150M at $1.45B valuation to scale Physical AI chip platform
- Targets embedded AI for robots, drones, and autonomous vehicles
- Next-gen chip aiming for 1,000 tera operations per second by 2028
Market signal
The substantial funding round highlights strong investor confidence in the growth potential of physical AI devices—hardware systems embedding AI for real-world perception and autonomous action. As robotics, drones, and autonomous vehicles proliferate, demand increases for efficient, effective AI compute engines that can operate with lower power and greater agility than traditional cloud-based AI solutions.
SiMa.ai’s focus on purpose-built silicon and integrated software reflects a market need for specialized AI platforms tailored to physical devices rather than repurposed cloud architectures. This positions the startup well to capture a significant share of the emerging multi-trillion-dollar physical AI market, providing alternatives to incumbents like Nvidia, especially in use cases where power efficiency and rapid deployment matter.
Operator impact
Operators and buyers of embedded AI hardware in robotics, automotive, and drone sectors can expect more competitive and efficient chip solutions coming to market. SiMa.ai’s emphasis on lower-power operation and faster deployment cycles may reduce integration complexity and operational costs, enabling faster iteration on AI-powered machine capabilities.
The development environment Palette Neat will help operators streamline software development for physical AI devices, potentially accelerating time-to-market for new autonomous features. Additionally, the planned Modalix system-on-chip aims to bring powerful AI compute performance suitable for mid- to high-end robotics and advanced driver assistance systems, expanding the options for hardware selection beyond established high-cost offerings.
What to watch next
Watch for SiMa.ai’s progress in delivering its next-generation chip, Modalix, expected to achieve 1,000 tera operations per second by mid-2028. Achieving performance and power efficiency metrics that rival or surpass Nvidia’s offerings will be critical to gaining broader adoption in key physical AI device segments.
Keep an eye on partnerships and customer wins as the company moves from development to deployment. Expansion into automotive AI cockpits, humanoid robotics, and advanced drones will determine how SiMa.ai’s platform scales across verticals. Monitoring competitor responses and ecosystem support for SiMa.ai’s software environment will also indicate the startup’s potential to disrupt the embedded AI compute market.