Nvidia CEO predicts Kazakhstan’s role as AI hub for Europe

Kazakhstan could become a major hub for artificial intelligence computing infrastructure serving Europe. This was stated by the founder and CEO of Nvidia Jensen Huang during his first visit to the country, reports CNBC.

At a public conversation with Deputy Prime Minister and Minister of Artificial Intelligence and Digital Development of Kazakhstan Zhaslan Madiev in Astana, Huang pointed to the country’s energy resources and plans to expand the capacity of data centers.

One of the key infrastructure projects of Kazakhstan was the Data Center Valley in Ekibastuz. According to CNBC, 300 MW of power capacity is already available, which is planned to be increased to 1 GW. The country attracts international developers of data centers with the cost of electricity, land availability and investment incentives.

In June, the Kazakh government announced plans to invest $10 billion in AI infrastructure together with US cloud company Firebird. The project provides for the creation of a computing cluster based on 100 thousand modern graphics processors, including Nvidia chips.

According to Nvidia, Firebird has received a US export license to supply 45,000 Nvidia Grace Blackwell GPUs to the country. The first stage of the project is being built together with Kazakhtelecom at the Data Center Valley site. Its commercial launch is scheduled for the third quarter of 2027.

The second phase can bring the total capacity of Firebird AI infrastructure in Kazakhstan to about 500 MW. The company also intends to establish Firebird Labs Kazakhstan in the country to support research, startups and application development.

Companies are considering placing computing power for artificial intelligence not only on Earth but also in space. In February, when announcing the acquisition of xAI by SpaceX, Elon Musk called the launch of AI satellites a priority for the combined company. In two to three years, space could become the cheapest place for such calculations. The entrepreneur explained the plans by the fact that the ground infrastructure will not be able to meet the growing demand of AI for electricity without harming the population and the environment.

In September, Google unveiled the Suncatcher orbital data center project. The company intends to explore the possibility of creating a scalable machine learning infrastructure in space. In low Earth orbit, satellites will be able to receive almost constant sunlight. Chip cooling remains a technical issue, with Google testing a combination of heat pipes and radiators, but the system’s performance in orbit has yet to be tested.