GMI Cloud

GMI Cloud announced record growth in 2026, reporting that live ARR from production GPU capacity grew 2.4x in the first half over year-end 2025, while contracted ARR from signed customer commitments surpassed $500 million — more than 8x its year-end 2025 ARR. The company also announced a sovereign AI infrastructure init As part of its global expansion, GMI Cloud has extended its data center footprint from the United States into Taiwan and Thailand, with further Asia-Pacific expansion underway. The international rollout is supported by partnerships with Macnica, Compal, and Taiwan Mobile, positioning GMI Cloud as a scaled AI-native clo

GMI Cloud announced support for NVIDIA's Vera Rubin platform following GTC 2026 Taipei, positioning itself as an inference-native cloud built for agentic AI workloads. The platform is designed for high-throughput, low-latency inference, multimodal model deployment across text, image, video, and audio, and long-context The GMI Cloud stack on NVIDIA infrastructure brings together high-performance AI infrastructure for training, inference, and production deployment; Prime Inference for optimized low-latency model serving; MaaS APIs unifying proprietary and open-source model access; dedicated enterprise endpoints; and orchestration laye

GMI Cloud announced support for NVIDIA's Vera Rubin platform at GTC 2026 Taipei on June 3, 2026, positioning itself as an AI-native cloud infrastructure provider built for production AI and agentic workloads. The company described its platform as inference-native, designed to help AI builders deploy, scale, and operate The platform combines several components: high-performance AI infrastructure for training, inference, and production deployment; Prime Inference for optimized low-latency model serving; MaaS APIs providing unified access to proprietary and open-source models; Dedicated Endpoints for enterprise-grade production inferenc