SiMa.ai introduces MLSoC Modalix, a GenAI-powered platform that enhances Edge AI performance and efficiency across multiple industries.
SiMa.ai has expanded its ONE Platform for Edge AI with the release of MLSoC Modalix, a GenAI-enabled solution. Designed for edge AI applications, MLSoC Modalix offers powerful configurations that allow developers to run complex AI models, including deep neural networks (DNNs), transformers, and GenAI, on a single chip. This technology addresses key challenges in edge computing, such as balancing performance and power efficiency, while providing faster, more secure insights across industries like defense, agriculture, medicine, and logistics.
One challenge in edge AI is handling multi-modal data (text, image, video) while maintaining low power usage. MLSoC Modalix addresses this by offering a flexible architecture that enables developers to deploy GenAI applications with minimal hardware and power requirements. Available in multiple configurations (M25, M50, M100, M200), the platform supports a wide range of models, including LLMs (Large Language Models) and LMMs (Large Multi-Modal Models), while offering seamless compatibility with existing SiMa.ai systems.
The platform leverages SiMa.ai’s patented software, ensuring efficient data handling and high performance without sacrificing power. Features like the integrated image signal processor (ISP), PCIe Gen 5 support, and eight Arm Cortex-A65 CPUs allow the chip to handle complex tasks in real-time. This makes it ideal for applications in smart vision, aerospace, retail, and industrial inspection. The system supports high-efficiency inferencing for both CNNs and transformer models, including Llama2, delivering substantial improvements in speed and accuracy at the edge.
With MLSoC Modalix, SiMa.ai provides a scalable solution that allows organizations to integrate GenAI into their operations while reducing total cost of ownership. The platform’s software-centric design ensures seamless upgrades and flexibility for future AI models, giving companies an efficient way to adopt cutting-edge AI technologies.