HUMAIN, a PIF company, is expanding its strategic partnership with NVIDIA to deploy up to 600,000 NVIDIA GPUs in Saudi Arabia and the United States over the next three years. This expansion includes establishing NVIDIA-powered data centers and utilizing NVIDIA Nemotron open technologies to train HUMAIN Chat, designed to serve the world’s more than 400 million Arabic speakers. Furthermore, HUMAIN will leverage NVIDIA Omniverse libraries to accelerate physical AI initiatives, building on a foundation established during its launch in May. These developments aim to advance sovereign AI infrastructure and deliver global-scale compute capability.
Expanding AI Infrastructure in US and Saudi Arabia
HUMAIN is significantly expanding AI infrastructure in both the U.S. and Saudi Arabia through a deepened partnership with NVIDIA. Plans include deploying up to 600,000 NVIDIA GPUs across both countries over the next three years, utilizing the latest NVIDIA GB300 platforms and Quantum-X800 InfiniBand networking. This expansion establishes new NVIDIA-powered data centers in Saudi Arabia and expands operations to AI data centers in the U.S., aiming to create globally competitive infrastructure for AI innovation.
A key component of the expansion involves multiple partnerships. HUMAIN will collaborate with xAI to develop a 500MW+ data center in Saudi Arabia, featuring approximately 18,000 NVIDIA GB300 GPUs, marking xAI’s first major deployment outside the U.S. Additionally, a new partnership with Global AI will focus on large-scale AI data centers in the U.S. and HUMAIN is collaborating with AWS to deploy up to 150,000 NVIDIA GPUs within a dedicated “AI Zone” in Riyadh.
This infrastructure expansion isn’t solely focused on compute power; HUMAIN also intends to leverage NVIDIA technologies to advance Arabic language AI. Specifically, they will utilize NVIDIA Nemotron open technologies to train HUMAIN Chat, an Arabic conversational AI app aimed at serving the world’s 400 million+ Arabic speakers. Furthermore, the partnership includes building advanced digital twin ecosystems for applications spanning energy, manufacturing, and smart cities.
Partnerships with Global AI, xAI, and AWS
HUMAIN is expanding its AI infrastructure through strategic partnerships with Global AI, xAI, and AWS. A new partnership with Global AI will focus on developing large-scale AI data centers and compute capacity in the United States, powered by NVIDIA technologies including the GB300 and Quantum-X800 InfiniBand networking. These U.S. facilities are designed for high-density compute, enabling large-scale model training and secure inference operations, bolstering a sovereign-cloud integration for enterprises and developers.
A parallel agreement with xAI will establish a network of world-class data centers in Saudi Arabia, anchored by a 500 MW+ facility. This deployment represents xAI’s first major expansion outside the U.S., initially utilizing approximately 18,000 NVIDIA GB300 GPUs to support the training of future Grok models and inference operations. HUMAIN is also collaborating with AWS, who will deploy and manage up to 150,000 NVIDIA GPUs within a dedicated “AI Zone” in Riyadh.
These partnerships collectively aim to strengthen HUMAIN’s position in global AI infrastructure. AWS will focus on supporting cutting-edge AI training and inference workloads with NVIDIA GPUs, while the combined efforts with Global AI and xAI establish significant compute capacity in both the U.S. and Saudi Arabia. This expansion is intended to secure a strategic foothold and fuel future AI innovation across multiple sectors.
We are working together to advance global AI infrastructure. By expanding our compute capacity in both the U.S. and Saudi Arabia with the latest NVIDIA AI infrastructure, HUMAIN, and our partners at Global AI, xAI, and AWS, secure an even stronger strategic foothold.
Tareq Amin, CEO of HUMAIN
Advancing Arabic AI Models and Digital Twins
HUMAIN is advancing Arabic AI models through the development of HUMAIN Chat, a conversational AI application. Utilizing NVIDIA Nemotron open technologies, the app is designed to serve over 400 million Arabic speakers, enabling creation, learning, and connection in their native language. This initiative is powered by HUMAIN’s ALLAM large language model and represents a key component of their strategy to deliver advanced, localized AI solutions.
The partnership with NVIDIA also focuses on building a leading Connected Digital Twin ecosystem. HUMAIN will integrate NVIDIA Omniverse libraries with its national AI stack, combining advanced compute infrastructure, foundation models, and AI agents with NVIDIA’s digital-twin and simulation technologies. This aims to power next-generation capabilities across sectors like energy, manufacturing, and smart cities, supporting large-scale infrastructure development.
HUMAIN plans to deploy up to 600,000 NVIDIA GPUs in Saudi Arabia and the US over the next three years. This includes a 500 MW+ data center in Saudi Arabia with approximately 18,000 NVIDIA GB300 GPUs, supporting the training of future models and serving as an inference outpost. Additionally, AWS will deploy up to 150,000 NVIDIA GPUs within a dedicated “AI Zone” in Riyadh, focusing on cutting-edge AI workloads.
