Kazakhstan as an AI Computing Hub: Timur Turlov’s Projects and Nvidia CEO’s Vision

Up to 300 servers powered by Nvidia B300 chips, a public GPU cloud, and a proprietary language model — these were among the plans Timur Turlov shared with Nvidia CEO Jensen Huang during Huang’s first visit to Kazakhstan. At a public event, the Nvidia chief discussed how the country could harness AI in industry, science, and agriculture.

First Visit to Kazakhstan

Timur Turlov, founder and CEO of Freedom Holding Corp., met with Nvidia CEO Jensen Huang during the latter’s first visit to Kazakhstan. Together with Zhaslan Madiyev, Deputy Prime Minister and Minister of Artificial Intelligence and Digital Development, they visited the Alem.ai International Center for Artificial Intelligence in Astana.

Turlov discussed the construction of Akashi Data Center and plans to launch a public GPU cloud for businesses and developers in 2027. He also introduced Huang to his company’s ecosystem, which encompasses banking, brokerage and insurance services, telecommunications, and everyday consumer services ranging from ticket purchases to grocery delivery.

One of the key topics was AI infrastructure. According to Turlov, Freedom is preparing to place an order for its first computing cluster based on Nvidia B300 chips. The cluster will comprise up to 300 servers, with investments estimated at hundreds of millions of dollars.

The cluster is expected to support the launch of a public GPU cloud in 2027. The service will allow companies and developers to rent GPU computing capacity to train and run AI models without having to purchase expensive hardware of their own.

Turlov also noted that Kazakhstan already has teams that have developed technologies for e-commerce and financial services. In his view, combining this expertise with access to domestic computing infrastructure could become one of the country’s competitive advantages.

A Sovereign AI Hub Worth Approximately $2 Billion

Freedom’s cooperation with Nvidia is developing as part of a sovereign AI hub project. The holding company announced an agreement with Kazakhstan’s Ministry of Artificial Intelligence and Digital Development in November 2025. The announced investment is approximately $2 billion.

Freedom will serve as the project’s principal financing and implementation partner, as well as the future operator of the hub. The ministry is expected to help create the conditions necessary for deploying and operating the infrastructure and to support workforce development programs.

The hub is planned for a site with 100 MW of available electrical capacity to power servers, cooling systems, and other equipment.

The project aims to establish computing capacity in Kazakhstan for the development and deployment of artificial intelligence. In this context, the term “sovereign” refers to the ability to develop domestic AI systems and maintain control over data processing within the country.

Akashi Data Center: Scale and Objectives

During the meeting, Turlov presented Akashi Data Center to Huang. The project is a data center campus under development in Astana, financed by Turlov through his personal investments.

The project is being developed through his private holding company, Turlov Private Holding, and is not part of Freedom Holding Corp. Olga Baskakova, CEO of Turlov Private Holding, previously clarified the distinction between Turlov’s private assets and the publicly traded holding company.

The campus will feature four buildings, more than 4,000 server racks — cabinets designed to house servers and networking equipment — and the potential to scale up to 100 MW. It is intended to accommodate computing infrastructure, data storage, cloud platforms, and AI services.

Akashi is being built to the Tier IV standard, which requires a high level of fault tolerance: the infrastructure must remain operational even if individual components fail.

According to Baskakova, Turlov Private Holding’s total investment in infrastructure development and the project’s launch will amount to at least $500 million.

Huang Sees Kazakhstan’s Potential as an AI Computing Hub

During a public discussion with Zhaslan Madiyev, Jensen Huang highlighted Kazakhstan’s potential to become an AI computing hub serving Europe. He linked this opportunity to the development of computing capacity, energy infrastructure, and international telecommunications connections.

The Nvidia CEO placed particular emphasis on physical AI — systems capable of analyzing the physical world and helping take action within it. According to Huang, this field encompasses a much broader range of applications than robotics alone.

For Kazakhstan, Huang identified mining, geology, agriculture, and materials science as particularly promising areas for AI adoption. These technologies could help explore mineral deposits, determine extraction methods, improve raw material processing, and develop new materials.

In his assessment, physical AI is still in the early stages of development, giving Kazakhstani companies an opportunity to help shape the emerging industry.

Ambition and People Matter More Than the Tools Themselves

Responding to a question about the future of professions, Huang suggested viewing artificial intelligence as a means of expanding human capabilities.

“If you have ambition, AI and robotics will help you reach the next level. They will enable you to do more,” he said.

As examples, the Nvidia CEO mentioned a farmer who could expand operations with the help of robots, a scientist able to explore more research areas using an AI assistant, and an engineer empowered to develop better solutions.

A Proprietary Language Model and Data Sovereignty

Turlov told Huang that Freedom had begun developing its own language model. The model is intended to reflect local context and cultural characteristics while meeting data security requirements.

According to Turlov, proprietary technology makes it possible to tailor the model to specific applications and retain control over how information is stored and processed.

During the public event, Huang separately emphasized that security must be built into AI systems from the outset. In particular, autonomous AI agents capable of working with files and software tools should be granted only the permissions necessary to complete their tasks, while their actions must remain subject to continuous oversight.

Asked what young people should study over the next five years, the Nvidia CEO recommended learning to work with artificial intelligence while also understanding how it functions, including the technology’s underlying principles, strengths, and limitations.