Recently, edge computing chip and software startup SiMa.ai announced the launch of its new product — the MLSoC Modalix chip, further solidifying its position in the edge AI market.
SiMa.ai has just completed a $70 million funding round, with investors including industry heavyweights such as Dell Technologies Capital.
The newly introduced Modalix chip features a 6-nanometer process, significantly improving in size and power consumption compared to its previous 16-nanometer counterpart. This chip is based on SiMa.ai's ONE platform, supporting edge AI, designed to handle advanced AI models including Convolutional Neural Networks (CNNs), Transformers, and Generative AI, while ensuring industry-leading performance and energy efficiency.
In a video interview, SiMa.ai's CEO Krishna Rangasayee expressed excitement about the Modalix launch. He noted that the chip is well-suited for industrial automation, healthcare, intelligent vision systems, aerospace, and defense, benefiting any scenario requiring multimodal processing.
Rangasayee also mentioned that the vision for the Modalix chip is to enable every device to possess human-like capabilities, with future devices being able to converse, express, and visualize. Meanwhile, Generative AI is gradually transforming the operational methods of various industries, although most current Generative AI is still primarily concentrated on desktop PCs and mobile devices.
SiMa.ai hopes to push these advanced technologies to dedicated devices like field robots and drones through Modalix.
The new platform is designed to handle multimodal AI tasks, capable of processing multiple inputs such as text, images, and audio simultaneously, and can run Meta's Llama2-7B parameter model on the chip. This capability holds significant potential for edge AI inference applications.
The performance configuration range of the Modalix series chips spans from 25 to 200 TOPS, designed specifically for high-load AI workloads, and excels in power reduction. Rangasayee stated that Modalix allows customers to no longer worry about power consumption and cooling issues, enabling easy implementation of low-power, high-performance AI applications.
Other companies in the industry have also recognized the potential of Modalix, with Elementary's CEO Arye Barnehama praising its efficiency and low power consumption. Additionally, SiMa.ai's Palette Edgematic software stack greatly simplifies AI deployment, making it accessible for more non-professional developers.
As the edge AI market continues to expand, analysts predict that the global edge computing market will double in growth over the next few years. In this context, SiMa.ai's Modalix chip demonstrates strong competitiveness, offering an attractive option for industries looking to leverage AI on edge devices.
Key Points:
🌟 SiMa.ai introduces the new MLSoC Modalix chip with a 6-nanometer process, significantly enhancing energy efficiency.
🤖 Modalix supports multimodal AI processing, suitable for various fields including industrial automation and healthcare.
⚡ The new chip design aims to simplify AI deployment, enabling non-professional developers to easily achieve high-performance applications.