INFINITIX and Advantech Expand Partnership to Boost GPU Efficiency for Industrial AI

As generative AI and real-time inference scale across industrial and enterprise environments, organizations are investing heavily in GPU infrastructure. However, many deployments struggle with low GPU utilization, limiting performance gains and return on investment.

INFINITIX, a leading provider of AI infrastructure orchestration software, today announced an expanded strategic partnership with Advantech, a global leader in industrial computing and edge AI. The collaboration delivers enterprise-grade and edge AI infrastructure solutions that significantly improve GPU efficiency while maintaining industrial-level reliability.

At the core of the solution is INFINITIX’s AI-Stack CTAs (Core Type Aware Scheduler), a software-defined scheduling technology designed to address GPU resource fragmentation. Modern GPUs combine CUDA Cores for general-purpose computing and Tensor Cores for AI acceleration, yet mixed workloads are typically executed sequentially, leaving large portions of compute capacity idle.

AI-Stack CTAs intelligently identifies GPU core types at runtime and enables CUDA-intensive and Tensor-intensive workloads to run concurrently on a single GPU. Through Kubernetes-native integration and transparent runtime interception, the technology requires no application changes and increases effective GPU utilization to over 80%, unlocking nearly double the usable compute performance.

As higher utilization places greater demands on system stability, Advantech provides industrial-grade GPU server platforms validated for sustained high-load operation. This includes the SKY-602E3 tower GPU server, designed for edge AI environments where reliability and continuous operation are critical.

By integrating AI-Stack CTAs across Advantech’s industrial edge and data center server portfolio, enterprises can deploy high-performance, reliable AI infrastructure across both centralized and distributed environments. The joint solution supports mission-critical AI workloads in manufacturing, telecommunications, healthcare, and other always-on industrial sectors.

Allen Chen, Chief Business Officer of INFINITIX, said the collaboration enables enterprises to “maximize GPU efficiency without sacrificing reliability,” helping organizations accelerate AI adoption and move from experimentation to large-scale production.