en.Wedoany.com Reported - Saudi Arabian AI infrastructure company Think is expanding its manufacturing and deployment capabilities for high-density liquid-cooled multi-GPU computing nodes, while continuing to develop its bare-metal orchestration software, ILM. The company plans to accelerate product production, team expansion, and international market deployment. After completing more projects in Saudi Arabia, it will extend its operations to other markets within the Gulf Cooperation Council.
Think's infrastructure solution in Saudi Arabia consists of AI Node computing hardware and the ILM software orchestration layer. The AI Node integrates multiple GPUs into high-density computing nodes and uses a liquid cooling system to control operating temperatures, enabling the computing equipment to continuously handle model training and inference tasks within a compact space. The entire platform can be deployed in data centers, offices, laboratories, and edge computing environments, not limited to large centralized server rooms.
ILM is responsible for unified management of servers, GPUs, and artificial intelligence tasks. After an enterprise deploys the model, the platform allocates computing resources based on task requirements, reducing GPU waiting time, idle time, and redundant configuration. Production environment test results released by Saudi Think show that the platform can maintain a GPU utilization rate exceeding 90%, higher than the industry common level of 30% to 50%. Higher utilization means the same set of computing equipment can handle more training and inference tasks, thereby reducing the need for additional servers.
This platform can operate using commercially available general-purpose GPUs without being tied to proprietary inference hardware. Saudi Think also plans to support the integration of GPUs from different vendors and specialized inference chips, allowing enterprises to configure computing clusters based on model type, computing performance, and existing equipment conditions, avoiding reliance on a single hardware ecosystem for the entire infrastructure.
Saudi Think is advancing multiple proof-of-concept projects, production deployments, and technical collaborations in Saudi Arabia, and is participating in local AI infrastructure initiatives. These projects focus on validating liquid-cooled computing nodes, bare-metal resource scheduling, and sovereign AI deployment capabilities, enabling models, data, and computing resources to operate within local infrastructure and reducing dependence on overseas cloud platforms.
The next phase of physical progress will focus on mass production of AI Nodes, assembly of liquid cooling systems, delivery of computing clusters, and independent development of the ILM platform. The company plans to expand its deployment scope within the Gulf Cooperation Council region over the next 18 months and gradually develop ILM into a software platform that can be independently integrated into different data centers and computing clusters.
