SiMa.ai Launches Next-Gen Physical AI Platform with Modalix™ MLSoC, SoM, and LLiMa

SiMa.ai, a pioneer in Physical AI solutions, has officially launched its next-generation Modalix™ Machine Learning System-on-Chip (MLSoC™), along with a new System-on-Module (SoM) and LLiMa™ software framework. These advancements mark a major leap forward in deploying LLMs, GenAI, and edge AI applications with high efficiency, low power consumption, and industry-wide scalability.

Modalix™: Powering the Next Era of Physical AI

Built on TSMC’s advanced N6 process and designed with Synopsys’ AI-powered EDA tools, Modalix™ delivers exceptional performance under 10 watts while supporting LLMs, transformers, CNNs, and GenAI workloads. Its flexible Arm-based architecture integrates seamlessly with real-time perception, decision-making, and natural language interaction for next-generation AI systems.

SoM & DevKit: Seamless Integration for Developers

The new Modalix SoM, co-developed with Enclustra, is pin- and form-factor compatible with leading GPU SoMs—offering a drop-in replacement for developers. Available with 8GB ($349) and 32GB ($599) options, the SoM integrates MIPI, memory, and I/O interfaces in a compact, power-efficient package.

A full DevKit ($1,499) enables plug-and-play prototyping, complete with integrated cooling, SSD, and camera support, empowering developers to move from concept to production with ease.

LLiMa™: Deploying LLMs at the Edge

The newly introduced LLiMa framework allows LLMs, LMMs, and VLMs to run fully on-device with zero cloud dependency. Featuring a curated model zoo, automated optimization, RAG support, and agent-to-agent capabilities, LLiMa unlocks on-device GenAI for robotics, smart vision, defense, healthcare, and industrial automation.

Industry Collaboration & Endorsements

• Arm praised Modalix for delivering power-efficient GenAI at the edge.
• Synopsys highlighted its role in enabling a successful first tapeout.
• TSMC confirmed its partnership in ensuring energy efficiency and reliability.
• Enclustra positioned the SoM as a ready-to-deploy AI platform for robotics, smart vision, and healthcare.

About SiMa.ai

Headquartered in San Jose, California, SiMa.ai is a leader in Physical AI, focused on scaling edge AI applications across industries including robotics, healthcare, automotive, aerospace & defense, and industrial automation. Learn more at www.sima.ai.