Energy and computing — the primary asset of the future
If the economic influence of states was previously determined by raw material reserves, sea access, and industrialization levels, the new key resource is now computing capacity and infrastructure for artificial intelligence (AI) projects. Growing computing capacity is elevating data centers to the rank of primary elements of competitiveness and national security.
The growing demand for digital capacity is also intensifying pressure on the global energy system. According to IEA forecasts, global data center electricity consumption will double by 2030, reaching 945 TWh — an amount exceeding the current total consumption of all of Japan. Electricity is ceasing to be merely a line item in business operating expenses. Guaranteed access to available capacity is now an independent strategic asset around which the logic of locating new digital facilities is organized.
“The situation is actually such that we stand on the threshold of a total reinvention of all global infrastructure. In a few years, having capacity for AI will influence macroeconomics no less than having raw materials or transport corridors. If data centers were once perceived by business as digital ‘warehouses’ for backups and archive storage, today they are the foundation — the new oil wells of the digital world. AI needs not just software; it needs physical, tangible things: concrete, heavy iron, and above all, massive, uninterrupted volumes of electricity. That is why new infrastructure projects are increasingly designed from the outset at scales that seemed excessive just a few years ago. For example, the total projected capacity of AKASHI in Astana is 100 MW,” notes Vladislav Minkevich, head of the AKASHI DataCenter being built in Astana.
A generational technology shift
AI technology has moved beyond BigTech and is becoming an infrastructure need for core industries — manufacturing, medicine, telecoms, and the public sector. According to McKinsey’s review, 62% of large organizations are already deploying autonomous AI agents. This is yielding real dividends: PwC data shows that a third of global company CEOs report that Generative AI use has directly driven revenue growth of 32% and business profitability growth of 34%.
The explosive growth of corporate demand for AI technologies has collided with a harsh physical reality: existing IT infrastructure is not ready for the new types of extreme workloads. If the industry standard just a few years ago was 5–10 kW per server rack, modern AI clusters demand 80–100 kW and above. A deep technological gap has emerged in the global market. Currently, fewer than 3% of active data centers worldwide meet the highest international fault tolerance standard, Tier IV.
High heat density on AI racks makes traditional air cooling systems completely obsolete. According to Uptime Institute, more than half of data center operators deploying AI are already rebuilding their facilities. Specifically, 52% of respondents are upgrading power distribution systems and 51% are switching to liquid server cooling. A global and quite painful technology transition has effectively begun.
“The vast majority of old data centers worldwide will not survive this transition. Modern AI chips in old halls is like trying to put a rocket engine on a cart — they simply cannot physically function there due to overheating. Fundamentally new buildings are needed. To grasp the scale of the loads: all of Astana consumes around 850 MW, while a single modern Tier IV digital facility at peak will consume 100 MW at one point. That is the scale of a huge factory — but a digital factory. Without deep re-engineering, old facilities will turn into unprofitable ‘IT barns’. The AKASHI data center is being built for new tasks — the project includes a separate AI-ready module of approximately 40 MW, designed for high-density next-generation computing workloads,” emphasizes Vladislav Minkevich.
Kazakhstan’s window of opportunity
The infrastructure crisis is compounded by the fact that traditional European IT hubs — Frankfurt, London, Amsterdam, Paris, and Dublin (the FLAP-D markets) — are experiencing a deficit of grid capacity. In Dublin, data centers are already approaching 80% of local electricity consumption, while in Amsterdam, London, and Frankfurt their share fluctuates around 30–40% of local demand. As a result, queues for connecting new facilities to energy grids have stretched to 7–10 years. Consequently, the market has stopped valuing data centers by floor space — guaranteed megawatts of grid capacity, not square meters, have become the key measure of value.
This situation creates a window of opportunity for developing regions. International companies are increasingly looking at alternative infrastructure geographies. In Minkevich’s view, Central Asia, and Kazakhstan in particular, has all the prerequisites to seize the initiative. The combination of available surplus energy, a cold climate that reduces cooling costs, a favorable geopolitical position at the junction of macro-regions, and political stability makes our country an ideal safe harbor for new digital giants. According to AKASHI DataCenter data, the first block of the project under construction is more than 103% reserved, with a significant portion of demand coming from foreign players considering Kazakhstan as an alternative location for digital infrastructure.
“Singapore, Switzerland, and Liechtenstein built their global competitiveness not on raw material extraction but on creating ‘infrastructure of trust.’ Today financial trust is being replaced by digital trust. The AKASHI project in Astana is being built to Tier IV requirements — the highest fault tolerance level in the Uptime Institute classification. Our project in Astana is an attempt to build the very ‘iron rails’ on which the economy of the future will run — a kind of exoskeleton for domestic business. The window of opportunity is open right now, and Kazakhstan must certainly take its rightful place in this story,” concludes AKASHI DataCenter head Vladislav Minkevich.
Source: Kazakhstanskaya Pravda