ASRock Industrial today announced a strategic collaboration with Axelera AI to deliver high-performance, power-efficient, and cost-effective AI inference solutions for edge computing. The collaboration integrates Axelera Metis AIPU cards into ASRock Industrial's edge AI platforms, delivering a compact, scalable AI inference solution optimized for real-world edge environments. The solution enables parallel AI processing by extending the system's native AI computing capabilities with up to 214 TOPS of additional inference through the Metis M.2 or PCIe AI acceleration cards. Powered by Metis AIPU, it delivers superior performance while significantly lowering power consumption and total cost of ownership (TCO) compared with conventional GPU-based architectures. Designed for edge AI applications including LLM, machine vision inspection, robotics, smart retail and city environments, the solution enables organizations to deploy advanced AI workloads where data is generated while minimizing the system footprint, power requirements, and infrastructure costs.
Validated Edge AI Platforms Accelerate Deployment
These validated configurations provide customers with a scalable path to deploying advanced AI capabilities on compact edge systems. On ASRock Industrial's expandable Edge AIoT platforms, including the iEPF-11000S, iEPF-10000S, iEPF-9500S, iEPF-9040S, and iEPF-9040VS series, the Axelera Metis PCIe card enables scalable, industrial-grade AI processing for demanding edge applications. For compact edge systems such as the AI BOX-A395, NUC BOX-358H, NUC BOX-255H, and NUC BOX-155H, the Axelera Metis M.2 card delivers high-performance AI inference in a compact, power-efficient form factor. Supporting multiple AI workloads running in parallel, these solutions enable real-time analysis for machine vision, robotics, retail and hospitality, warehouse automation, smart surveillance, and intelligent transportation.