Advanced Micro Devices, Cisco (CSCO) and HUMAIN confirm plans to form Joint Venture to deliver AI infrastructure
1:33 PM ET, 11/19/2025 - Briefing.com
AMD, Cisco and HUMAIN to invest in a joint venture and serve as its exclusive technology partners, deepening their multi-year strategic collaboration announced in May during U.S. President Donald J. Trump's visit to the Kingdom of Saudi Arabia.The joint venture plans to deploy up to 1 GW of AI infrastructure by 2030, with the shared ambition to expand capacity to multiple gigawatts, as a key pillar of HUMAIN's overall ambitions.This joint venture is expected to begin operations in 2026, with a phase 1 deployment of 100 MW planned, powered by HUMAIN modern data center capacity, AMD Instinct MI450 Series GPUs and Cisco critical infrastructure solutions
.
.
.
.
NVIDIA: HUMAIN also confirms it is entering into a strategic partnership with American sovereign AI infrastructure company Global AI
1:03 PM ET, 11/19/2025 - Briefing.com
The strategic partnership between HUMAIN and Global AI - an NVIDIA Cloud Partner - seeks to accelerate the expansion of AI infrastructure globally. Together, the companies plan to deliver large-scale AI data center and compute capacity in the United States. The campus will be designed for high-density compute, supporting advanced AI workloads, large-scale model training, and secure inference operations. The off-premises, air-gapped architecture will include liquid-cooled NVIDIA AI infrastructure and comprehensive data-sovereignty controls, creating globally competitive infrastructure for enterprises, public-sector entities, and AI model developers.The facilities will provide AI infrastructure services for enterprise, public-sector, and AI-model-development partners, enabling model training, large-scale inference, and sovereign-cloud integration within locally governed ecosystems with access to the latest NVIDIA AI infrastructure, including NVIDIA Grace Blackwell Ultraplatforms connected with NVIDIA Quantum-X800 InfiniBand networking to be deployed as initial compute capacity. |
