SDAIA is thrilled to announce a groundbreaking collaboration with NVIDIA that will propel Saudi Arabia to the forefront of AI innovation.
At the heart of this collaboration is the integration of the latest NVIDIA technological advancements to enable developers to easily build and deploy AI applications with the ALLaM Arabic LLM model, ensuring unparalleled performance and scalability for region-specific workloads.
NVIDIA NeMo
Leveraging NVIDIA NeMo, part of the NVIDIA AI Enterprise software platform for large-scale language model training and NeMo Guardrails for AI safety, this strategic move will provide developers with a faster, more accessible path to building generative AI applications with ALLaM's capabilities and set a new standard for AI-driven language models in the region, making it a pivotal element in Arabic language AI.
Additionally, companies will have access to recently announced NVIDIA NIM Agent Blueprints to support the quick and simple development of customised AI agents for any industry.
NVIDIA Blackwell architecture
Plans are in place to establish one of the largest high-performance computing data centres
In addition to these advancements in AI applications, SDAIA is working closely with NVIDIA to significantly scale its supercomputing infrastructure. Plans are in place to establish one of the largest high-performance computing data centres in the MENA region by expanding SDAIA's existing NVIDIA DGX SuperPOD infrastructure.
This expansion is planned to integrate NVIDIA's most advanced technologies, including the upcoming NVIDIA Blackwell architecture, and is expected to eventually grow to over 5,000 GPUs, setting a new benchmark for digital innovation and infrastructure in Saudi Arabia.
LLM training and deployment
This partnership builds upon the successful deployment of the initial SDAIA NVIDIA DGX SuperPOD AI supercomputer and includes a roadmap to exponentially expand supercomputing capabilities while adopting NVIDIA's best practices for LLM training and deployment.
By integrating NVIDIA Triton Inference Server for efficient model deployment and NVIDIA TensorRT software for optimised inference, SDAIA is ensuring that its AI infrastructure remains at the cutting edge of global technology trends. Together, SDAIA and NVIDIA are driving a new era of AI-powered transformation in Saudi Arabia, reinforcing the Kingdom's leadership in AI and digital innovation on the global stage.
