SiMa.ai's Modalix Platform Wins Top Award, Signaling a Power-Efficient Future for Edge AI

SAN JOSE, CA – April 20, 2026 – SiMa.ai, a company at the forefront of embedding intelligence into physical systems, has secured a significant industry accolade for its latest technology. The firm’s Modalix™ MLSoC System-on-Module (SoM) was named the "Best Edge AI Board" in the prestigious 2026 Product of the Year Awards by the Edge AI and Vision Alliance, a leading industry partnership of over 100 technology companies.

The award recognizes the Modalix platform's innovative approach to bringing high-performance machine learning and generative AI capabilities to the network's edge—the real-world environment where data is generated. Launched in 2025, the Modalix SoM is engineered to deliver powerful AI processing while consuming less than 10 watts of power, a critical threshold for devices operating under tight thermal and energy constraints.

This recognition validates SiMa.ai’s focus on what it terms "Physical AI," a paradigm shift from passive data processing to enabling machines that can reason and act autonomously. "Being recognized by the Edge AI and Vision Alliance affirms what SiMa.ai was founded to deliver: high performance without the power drain, and a true platform for this transition," said Durga Peddireddy, Vice President of Product Management & Partnerships at SiMa.ai. "Our purpose-built Modalix MLSoC, paired with Palette, our software suite, addresses the full spectrum of use cases — from computer vision to reasoning-based analytics."

Redefining the Edge with Physical AI

The core of SiMa.ai's innovation lies in its ability to solve one of the biggest bottlenecks in deploying AI outside the data center: power consumption. Traditional AI hardware, particularly high-power GPUs, often proves impractical for edge applications due to excessive heat generation and energy requirements. The Modalix platform directly confronts this challenge.

The Modalix MLSoC (Machine Learning System-on-Chip) is a compact powerhouse built on a 6nm process. It integrates an 8-core Arm Cortex-A65 CPU, a Synopsys ARC EV74 vision processor, and a specialized neural network accelerator capable of 50 TOPS (Tera Operations Per Second). This architecture allows it to handle complex, multimodal AI workloads—processing text, images, and audio simultaneously—directly on the device. Impressively, the company has demonstrated that Modalix can run variants of large language models, such as Meta's Llama 2-7B, on-device, a feat that typically requires much more power-hungry hardware.

By enabling advanced perception and real-time decision-making locally, the platform allows customers to modernize existing systems without costly infrastructure overhauls. This is crucial for industries looking to add intelligence to legacy equipment, bringing powerful AI capabilities closer to the data source and enabling truly autonomous operations.

A Strategic Win in a Booming Market

Winning an award from the Edge AI and Vision Alliance, a respected organization that includes industry giants and innovators, provides more than just bragging rights; it offers crucial market validation. The Alliance's focus on practical, deployable AI technology means its awards carry significant weight among engineers, developers, and decision-makers.

This validation comes at a pivotal time. The global edge AI market, valued at over $24 billion in 2025, is on an explosive growth trajectory. Market research firms project it will surpass $100 billion by 2030, fueled by the proliferation of IoT devices and a surging demand for applications that require low-latency processing and enhanced data privacy. Industries from manufacturing and automotive to healthcare and retail are racing to adopt edge AI for faster, more secure, and more efficient operations.

Investor confidence in SiMa.ai's strategy appears strong. The company, founded in 2018, has successfully raised a total of $355 million in funding from prominent backers including Maverick Capital, Dell Technologies Capital, and Point72. This substantial capital is fueling its global expansion and accelerating the development of its Physical AI platform, positioning it to capture a significant share of this rapidly expanding market.

Challenging Incumbents on Power and Performance

SiMa.ai is entering a competitive arena dominated by established players like NVIDIA, Intel, and Qualcomm. However, the company is carving out a distinct niche by relentlessly focusing on power efficiency. While competitors often prioritize raw performance, SiMa.ai champions a "performance-per-watt" metric, claiming its technology is up to 10 times more efficient in real-world applications than its immediate rivals.

This focus on efficiency is not just a technical detail; it's a core strategic advantage. In a direct and telling move, SiMa.ai designed its Modalix SoM boards to be socket-compatible with some of NVIDIA's popular Jetson modules. This allows potential customers to directly swap out a competing board and test SiMa.ai's platform within their existing designs, creating a low-friction pathway to adoption and a direct challenge to the market leader.

By delivering a complete, software-centric platform that includes its Palette software suite and LLiMa framework for deploying on-device language models, SiMa.ai aims to simplify the notoriously complex process of deploying ML models at the edge. This combination of hardware efficiency and software accessibility is designed to lower the barrier to entry for companies seeking to build next-generation intelligent products.

From Theory to Reality: AI in Action

The practical implications of this technology are vast. Instead of abstract benchmarks, SiMa.ai has demonstrated tangible applications that showcase the Modalix platform's capabilities. In one compelling demonstration, a single Modalix-powered PCIe card was shown running AI video analytics on 80 separate video streams simultaneously. The system performed complex, multi-model tasks—including vehicle detection, person tracking, and pose estimation—all on-chip, without cloud dependency, and while consuming under 10 watts.

This level of on-device processing power unlocks scalable deployments in smart cities, automated factories, and intelligent warehouses, where hundreds of cameras require real-time analysis. The ability to perform these tasks without relying on a central server reduces latency, cuts bandwidth costs, and enhances security by keeping sensitive data local.

Across robotics, industrial automation, smart vision, and autonomous drones, the promise of Physical AI is becoming a reality. By providing the efficient computational backbone for these systems, platforms like Modalix are enabling a new class of devices that can perceive, reason, and interact with the world in real time, moving artificial intelligence from the cloud into the fabric of our physical environment.