AI/ML

SiMa.ai Unveils Modalix MLSoC Platform to Revolutionize Edge AI

by VARINDIA 2025-08-13

Powered by TSMC’s advanced N6 process, Modalix meets strict embedded power, thermal, and reliability requirements—setting a new benchmark for next-generation edge AI solutions.

SiMa.ai has officially launched its second-generation Machine Learning System-on-Chip (MLSoC) – Modalix – now in full production and available worldwide. Designed to overcome the biggest hurdles in edge AI—power efficiency, real-time performance, and easy integration—Modalix is set to transform industries including robotics, autonomous vehicles, industrial automation, aerospace, defense, smart vision, retail, and healthcare.

As organizations increasingly deploy AI at the edge, they face limitations with cloud-based AI, which often suffers from high latency, significant power consumption, and complex integration. Modalix solves these challenges by delivering industry-leading performance and accuracy under 10 watts, supporting advanced workloads such as Large Language Models (LLMs), transformers, Convolutional Neural Networks (CNNs), and Generative AI (GenAI) applications—all on-device.

Next-Gen Edge AI Architecture

Built on a powerful Arm-based architecture with a native GenAI software stack, Modalix™ enables real-time perception, decision-making, and natural language interaction. It supports key interfaces like camera, Ethernet, and PCIe, ensuring adaptability across diverse AI-driven systems.

In addition to Modalix, SiMa.ai introduced:

Pin-Compatible System-on-Module (SoM): Developed with Enclustra, this compact, power-efficient module is a direct replacement for popular GPU SoMs, integrating MIPI, memory, and essential I/O for faster time-to-market.
LLiMa Framework: A unified, offline platform for running LLMs, LMMs, and VLMs with features such as automated quantization, a curated model zoo, and agent-to-agent system support.
Developers benefit from the integrated Palette SDK, enabling rapid prototyping to production deployment.

Industry leaders including Arm, Synopsys, TSMC, and Enclustra praised the platform for its innovation in on-device AI and Physical AI systems. Krishna Rangasayee, CEO of SiMa.ai, emphasized that Modalix is accelerating global adoption of GenAI at the edge, making AI-powered devices smarter, faster, and more sustainable.