TAIPEI (Taiwan News) — Compal Electronics announced Thursday it has formed a partnership with Finland-based AI cloud provider Verda to accelerate AI infrastructure deployment in Europe and Asia.

Under the agreement, Compal will supply high-density liquid-cooled GPU server systems designed for large-scale AI training, inference, and agent-based applications, per CNA. The company said the platforms are built to improve cooling efficiency and energy performance for sustainable AI cloud deployment.

Verda operates high-density GPU data centers across Europe and provides computing resources for AI labs, research teams, and startups. Compal said the partnership will help expand advanced AI computing capacity as demand for generative AI services continues to rise.

Separately, Japanese software services provider Datasection Inc. announced an expanded partnership with Compal and plans to purchase 635 GPU servers using Nvidia’s latest B300 architecture. The investment is valued at about NT$10.2 billion (US$325 million).

Datasection said the servers will serve as the core infrastructure for its next-generation AI data center project. The company added that the systems are expected to provide large-scale computing capacity for major global cloud service providers.