M
MECC
Back to Insights

From Hydropower to Compute Power: Monetizing Mekong Megawatts

|MECC Research Unit

This article explores the untapped potential of Laos's hydropower exports in the context of AI-driven compute demand across Asia, outlining how regional energy infrastructure can evolve into compute infrastructure.

**By MECC Research Unit** January 2026

---

Executive Summary

Laos generates over 75% of its electricity from hydropower, exporting most of it to Thailand and Vietnam under legacy pricing agreements. Meanwhile, global demand for compute — particularly AI training and inference — is projected to exceed 1,000 TWh by 2030. The Mekong region, with its energy surplus and cooling capacity, is uniquely positioned to benefit.

---

Key Data Points

  • **Laos exports 80–90% of its generated electricity**, mostly to Thailand at rates as low as $0.05/kWh.
  • **Thailand's AI infrastructure growth** is projected at 19% CAGR, reaching 1.2 GW of compute demand by 2028.
  • **Singapore's AI hosting market** is supply-constrained, with land/cooling limits forecast to cap domestic buildouts by 2026.
  • **IDC estimates** global compute workloads will triple by 2030, requiring ~$1.3 trillion in new datacenter infrastructure.
  • **Energy cost is 35–60%** of total operating expense in hyperscale AI deployments.

    • ---

    Strategic Analysis

    The value of a kilowatt-hour (kWh) changes when it powers AI. At $0.05/kWh, Laos earns $50M/year from 1GW exports. That same electricity — routed to AI — could underpin $250M+ in compute revenue, with a portion flowing back as fees or equity.

    MECC's strategy is to:

  • Trigger price discovery by aligning energy with compute
  • Introduce the narrative of sovereign compute corridors
  • Simulate demand via regional interest (SG/JP)
  • Use scarcity optics to renegotiate existing power purchase agreements (PPAs)

    • ---

    Recommendations

    1. Laos should classify a portion of hydropower exports as "compute-eligible," setting differentiated rates 2. Regional leaders should convene around sovereign compute financing 3. Thailand should secure future compute capacity via bilateral renegotiation or risk regional loss 4. Multilateral lenders should treat compute as critical infrastructure and update funding frameworks accordingly

    ---

    Source Notes

  • Lao PDR Energy Statistics Yearbook (2024)
  • ASEAN+3 Power Grid Reports
  • Hyperscale AI Trends, McKinsey Global Institute (2025)
  • IEA: Global Electricity Outlook 2025