GMI Cloud Expands Support for NVIDIA Vera Rubin AI Factory Platform

GMI Cloud announced its support for the next stage of agentic AI factories built on the NVIDIA Vera Rubin platform at GTC 2026 in Taipei, according to a press release. The company said it is aligning with NVIDIA’s AI Factory ecosystem to deliver infrastructure optimized for large-scale inference, multimodal reasoning, and continuous AI operations.

The company is developing an inference native cloud platform designed to help AI teams deploy, scale, and operate production workloads across text, image, video, audio, and agentic workflows. The platform combines high-performance infrastructure, optimized model serving through its Prime Inference service, and unified MaaS APIs for access to proprietary and open-source models.

GMI Cloud is also adopting NVIDIA Confidential Computing to provide secure execution environments for enterprise AI systems handling proprietary or regulated data. The platform supports multi-tenant operations, dynamic scaling, and resource optimization to reduce token costs while maintaining performance.

CEO Alex Yeh said the company aims to enable developers to move from prototype to production more quickly by combining compute orchestration, inference delivery, and developer-friendly APIs. The collaboration with NVIDIA focuses on secure, efficient infrastructure to support the growing demand for autonomous and multimodal AI systems.