Compal Electronics said it is adding manufacturing capacity in Taiwan and the United States as it expands work with GPU cloud provider Verda, aiming to support the company’s deployments across multiple regions.
Compal said production in the United States could support deployments there, while capacity in Daxi, Taiwan, could help Verda expand across Europe, the Middle East, Africa and Asia. The announcement connects the companies’ roles across the infrastructure chain: Verda operates GPU cloud services and handles deployment, while Compal contributes server engineering and manufacturing.
Compal identified its SGX30-2 server as a system for the expanded collaboration. It said the platform is based on NVIDIA HGX B300 systems and includes eight NVIDIA Blackwell Ultra GPUs. That configuration describes the system, not a confirmed number of servers for Verda; Compal did not say how many units would be supplied or deployed.
Verda Chief Operating Officer Jorge Santos said the manufacturing footprint could help the company place computing capacity nearer to customers. “Compal gives us the manufacturing reach to bring capacity online closer to our customers,” Santos said. Compal cited rising demand for AI computing and the need to deploy infrastructure across regions as reasons for the expanded work. It said the wider production footprint is intended to improve deployment flexibility, supply resilience and speed.
The companies announced a strategic partnership on May 7, 2026. Compal said it would supply GPU server systems for Verda’s infrastructure in Europe and Asia-Pacific, describing those platforms as high-density and liquid-cooled.
Compal did not disclose an order size, installation schedule or financial terms. It also gave no revenue estimate or cost allocation, and announced no next deployment milestone. NVIDIA’s schedule lists the GTC Berlin conference and exhibits through Oct. 22; Verda listed booth 3059 in Hall 3 at STATION Berlin.
