TLDR: Cerebras Systems has launched ‘Cerebras for Nations,’ a new global initiative designed to assist governments worldwide in establishing, accelerating, and scaling their sovereign AI capabilities. The program focuses on co-designing AI supercomputers, co-developing advanced AI models, and deploying secure, high-performance AI infrastructure, leveraging Cerebras’ Wafer-Scale Engine technology.
Sunnyvale, CA – November 11, 2025 – Cerebras Systems, a leader in high-performance AI infrastructure, today announced the launch of ‘Cerebras for Nations,’ a groundbreaking global program aimed at empowering national governments to build, accelerate, and scale their sovereign artificial intelligence initiatives. This strategic move underscores Cerebras’ commitment to advancing AI capabilities on a global scale, ensuring nations can develop and control their own AI resources.
The ‘Cerebras for Nations’ initiative will see Cerebras collaborate closely with international partner governments and their respective private sector ecosystems, including datacenter, cloud, and AI companies. The program is structured around three core pillars designed to foster robust sovereign AI development.
Firstly, Cerebras will focus on the co-design and construction of world-class AI supercomputers. These advanced systems will be powered by Cerebras’ revolutionary third-generation Wafer-Scale Engine (WSE-3) chips and WSE-3 equipped CS-3 systems. These technologies are engineered to deliver industry-leading speeds for generative AI inference and AI model training, enabling what Cerebras describes as ‘GPU-impossible inference speeds’ for critical applications such as voice processing, complex reasoning, and sophisticated agentic workflows. Furthermore, these supercomputers will facilitate secure, domain- and language-specific state-of-the-art AI model training in a fraction of the time typically required by traditional GPU clusters. The initiative plans to deploy these sovereign systems either on-premises within participating countries or make the infrastructure accessible via secure cloud environments.
Secondly, the program emphasizes the co-development of state-of-the-art AI models. Cerebras for Nations will provide dedicated teams of its world-class AI research staff to partner with local government and industry project teams. This collaborative approach aims to rapidly develop new, cutting-edge models tailored for scientific research, industrial applications, and societal benefits. These models will then be optimized and deployed for production inference, ensuring complete infrastructure and data security. A key aspect of this pillar is that full ownership of the trained models will remain with the host institution, providing nations with autonomy over their AI assets.
Andy Hock, Chief Strategy Officer at Cerebras, expressed enthusiasm for the new program, stating, “We are thrilled to announce the launch of Cerebras for Nations, expanding access to AI compute and adding partnerships to accelerate AI for science and society around the world. This initiative emerges from the extraordinary work we’ve done for years with the U.S. government, in the UK, and with our strategic partners G42 in the UAE. We look forward to expanding those partnerships and adding many more.”
Cerebras Systems highlights its AI infrastructure as the ‘fastest AI infrastructure,’ leveraging the ‘World’s Fastest Processor,’ the Cerebras Wafer-Scale Engine, which is purpose-built for ultra-fast AI. The company claims its technology offers industry-leading speed, scale, and quality, with the potential for up to 30 times faster inference speeds compared to conventional GPU clouds, alongside reduced infrastructure costs. The systems are compatible with popular models such as OpenAI, Qwen, and Llama, and are designed to provide instant answers, prevent agent stalls, and accelerate code generation.
Customer testimonials underscore the practical benefits of Cerebras’ technology, citing improvements in enterprise search, real-time feature implementation, significantly faster cancer-drug response prediction, and accelerated drug discovery processes.
Also Read:
- Kazakhstan to Host $2 Billion Sovereign AI Hub, a Joint Venture by Freedom Holding Corp. and NVIDIA
- Cisco Introduces Unified Edge Platform for Local AI Processing
This initiative marks a significant step in democratizing access to advanced AI capabilities for national governments, fostering innovation, and ensuring data sovereignty in the rapidly evolving global AI landscape.
