TAIPEIMay 28, 2026 /PRNewswire/ — Compal Electronics (Compal; stock code 2324) today announced a collaboration with GMI Cloud, a Silicon Valley-based AI infrastructure service provider, to accelerate the deployment of next-generation AI inference and agentic AI application infrastructure. As part of this collaboration, GMI Cloud will adopt a high-performance GPU server platform to support the growing demands of large-scale AI training and inference workloads. This deployment will serve as a critical foundation for GMI Cloud’s continued expansion of AI-driven services and data-oriented applications.
Four individuals pose in front of a GMI logo, holding signed documents as part of a collaboration announcement between Compal and GMI Cloud
With the rapid growth of AI computing demands, cloud service providers are actively expanding their infrastructure to address increasingly complex application scenarios, including large language models, large-scale AI inference services, agentic AI systems, and real-time AI applications. Leveraging its expertise in high-density server design, advanced thermal architecture, and system integration, Compal helps customers build efficient and reliable large-scale deployment environments.
Chang Yao-wen, Vice President of Compal’s Infrastructure Solutions Business Group, stated: “As AI workloads rapidly shift toward large-scale inference and emerging agentic AI applications, infrastructure demands are evolving toward higher density, greater efficiency, and faster deployment. We are pleased to collaborate with GMI Cloud to build an infrastructure platform optimized for next-generation AI inference and practical AI application deployment.”
Alex Yeh, Founder and CEO of GMI Cloud, stated: “As AI moves from model experimentation to practical deployment, scalable inference infrastructure is becoming increasingly critical. Through this collaboration with Compal, we will continue to expand the infrastructure capabilities needed to support agentic AI workloads and large-scale inference services.”
In addition to infrastructure deployment, Compal and GMI Cloud will jointly showcase the latest collaboration results at COMPUTEX 2026. At Compal’s booth (M0804), visitors can experience GMI Cloud’s latest agentic AI and inference application scenarios, demonstrating how advanced infrastructure supports the practical deployment and implementation of next-generation AI services. Meanwhile, at GMI Cloud’s booth (R0302), Compal’s high-performance AI server platform, the SGX30-2, will be on display. This system is specifically designed to support the NVIDIA HGX B300 architecture, providing the high performance and scalability required for large-scale AI training and inference. Optimized for high-density deployment, advanced thermal management, and system-level integration, this platform offers a robust foundation for next-generation AI infrastructure.
This collaboration highlights Compal’s continued role in supporting emerging cloud service providers and AI infrastructure operators, helping them expand computing capabilities and accelerate service deployment. With its global manufacturing footprint and comprehensive supply chain, Compal is well-positioned to assist customers in efficiently deploying AI systems across various regions.
About Compal
Founded in 1984, Compal is a leading manufacturer of notebook computers and smart devices. With strong R&D and manufacturing capabilities, it has long collaborated with industry partners to create brand value. Compal has received numerous international design awards, was ranked among Taiwan’s top 7 manufacturing companies by CommonWealth Magazine in 2025, and has been listed in Forbes’ Global 2000 for many years. In recent years, it has actively expanded into emerging businesses such as cloud servers, automotive electronics, and smart healthcare. Headquartered in Taipei, Compal has manufacturing facilities in the United States, Taiwan, China, Vietnam, Mexico, Brazil, and Poland. For more information, please visit: www.compal.com
About GMI Cloud
GMI Cloud is a Silicon Valley-based AI infrastructure company that provides an “inference-first” full-stack GPU cloud and AI platform. By integrating high-performance computing power with software capabilities, we help enterprises and governments rapidly deploy and scale AI applications, while supporting sovereign AI initiatives to meet different national and regulatory requirements, building controllable and scalable AI capabilities.
For more information, please visit: gmicloud.ai