SK Telecom to build up to 15 GW AI data center capacity by 2035 with Nvidia
SK Telecom plans to scale its AI data center capacity to as much as 15 gigawatts by 2035, aiming to turn South Korea into a regional hub for AI computing. The first phase targets 5 GW, scheduled to begin operations in 2029.
The initiative is backed by a June 7, 2026 joint project between SK Group and Nvidia. SK Telecom will use Nvidia’s DSX platform to build gigawatt-scale AI cloud infrastructure. To execute the plan, SK Telecom set up a dedicated subsidiary, SK Hyper, in July 2026.
Funding details: the board approved an initial KRW 750 billion (about $506 million–$510 million) to be deployed by 2030. Cost estimates for a single 1 GW-class facility are around KRW 70 trillion. Early site work starts in Ulsan, with expansion plans for the Chungcheong and Honam regions.
In a June 30, 2026 6-K filing, SK Telecom laid out the full roadmap to reach 15 GW of AI data center capacity and indicated that additional strategic partners and customer contracts will help shape financing.
Key risks include execution and scale challenges: power generation and grid capacity, regulatory approvals, construction timelines, and the fact that the initial ~$506 million–$510 million commitment is only a fraction of total capital needs. Later-phase financing remains preliminary.
Neutral
This is a major AI infrastructure capex story (SK Telecom planning up to 15 GW of AI data center capacity by 2035 with Nvidia), but it has no direct linkage to specific crypto assets, tokens, or on-chain protocols in the article. As a result, the likely trading impact on crypto markets is indirect and limited.
Short term: Large corporate AI infrastructure announcements can briefly lift “tech/AI sentiment” in broader markets, but without any crypto-specific catalysts (no token launches, partnerships with crypto firms, or regulatory moves), traders typically do not reprice major coins materially.
Long term: If AI compute demand accelerates, it can support the broader tech ecosystem and may indirectly influence sectors that ultimately affect crypto narratives (e.g., infrastructure, data, energy). Still, execution risks (power, grid, permits, and financing for later phases) reduce certainty—similar to how past infrastructure-heavy announcements often translate into gradual, not immediate, market effects.
Overall, this reads as a strategic industrial investment rather than a crypto market driver, so a neutral impact rating is appropriate.