# Geothermal Energy's Role in Meeting Data Centers' Future Power Needs

> The future of AI is not just about advanced chips and talent; it's driven by a relentless need for electricity. Discover how the race for power is reshaping the landscape of artificial intelligence.

**Source**: datacenterdynamics.com | **Published**: 2026-09-06 | **Type**: article

## Key Facts

- US data centers' electricity demand could quadruple to 106GW by 2035, stressing grid capacity.
- Geothermal energy's capacity factor above 90% positions it as a strategic asset for data centers.
- Fervo Energy's $10.5B valuation highlights investor confidence in future clean energy revenue streams.
- Companies pivoting to "behind the meter" generation mitigate grid risks, enhancing competitive positioning.
- 200 geothermal campuses could meet 25% of projected US data-center consumption by 2030, reshaping supply.

## Summary

The race for artificial intelligence supremacy is increasingly defined by a critical resource: electricity. As companies invest heavily in AI, the demand for reliable, affordable, and clean power is surging. The U.S. data center sector, which consumed approximately 25 gigawatts (GW) in 2024, is projected to require over 106 GW by 2035, marking a more than fourfold increase in just a decade. This shift represents a significant challenge for energy infrastructure, as data centers could account for nearly 12% of U.S. electricity consumption by 2030.

The urgency of this situation stems from the economics of AI operations. High-performance computing clusters, which are essential for training and inference, must operate continuously to justify their substantial capital investments. Any interruptions in power supply can lead to significant losses. A single 100 MW data center can consume around 876 gigawatt-hours (GWh) annually, highlighting the scale of energy required to support AI workloads. However, the existing power grid struggles to keep pace with this escalating demand, leading hyperscale companies to explore alternative energy solutions, including on-site generation.

The current energy landscape is dominated by natural gas, which constitutes about 43% of the U.S. energy mix for data centers. While it remains the most flexible and politically favored option, it is also carbon-intensive. Other sources, such as nuclear and renewables, present their own challenges. Nuclear energy offers a reliable, low-emission alternative but faces long construction timelines. Wind and solar, while cost-effective, are intermittent and cannot guarantee continuous power. The emergence of enhanced geothermal systems (EGS) offers a promising solution. With the potential for high capacity factors and minimal emissions, geothermal energy could play a pivotal role in meeting the energy demands of the AI sector.

As data centers become some of the largest energy consumers, they are transforming into aggressive buyers in the energy market. Companies are signing long-term power purchase agreements and investing in new energy generation to ensure reliable supply. This shift reflects a broader strategy among hyperscalers to control their energy sources, insulating themselves from grid congestion and price volatility. The focus on reliability is changing site selection criteria, with capacity factors becoming a key metric for evaluating energy sources.

The implications for the market are significant. With many hyperscalers committing to carbon-free operations by 2030, there is a growing emphasis on the carbon intensity of energy sources. This trend is likely to drive up the value of clean energy solutions, as companies seek to align their operations with sustainability goals. Enhanced geothermal systems, in particular, are gaining traction, with capital flowing into projects aimed at delivering firm, carbon-free power directly to data centers.

The potential for geothermal energy is substantial. If it were to capture a quarter of the projected new data center load in the U.S. by the mid-2030s, it could lead to the establishment of approximately 200 dedicated geothermal campuses. Each of these facilities could generate significant amounts of energy, potentially supplying a substantial portion of the total projected demand for data centers.

For investors and industry stakeholders, the implications are clear: as the demand for AI power continues to grow, the ability to pair reliable, clean energy generation with data centers will become increasingly valuable. The market is signaling that infrastructure capable of delivering 24/7 carbon-free power will command premium economics, positioning it as a critical asset in the AI race. The strategic focus on energy reliability and sustainability will define competitive dynamics in the data center sector, ultimately determining which companies emerge as leaders in the evolving landscape of artificial intelligence.

## Entities

- **Companies**: Google, Fervo Energy, Meta, Sage Geosystems, XGS Energy, Ormat, Breakthrough Energy Ventures, CalSTRS
- **Technologies**: Enhanced Geothermal Systems, GPUs
- **Organizations**: Department of Energy, Lawrence Berkeley National Laboratory, International Energy Agency, NREL

## Key Concepts

AI power demand, data centers, electricity consumption, geothermal energy, power-purchase agreements, carbon-free power, capacity factor, hyperscalers

## Definitions

- **Enhanced Geothermal Systems**: A technology that utilizes directional drilling and reservoir engineering to extract heat from hot dry rock, providing a firm and carbon-free energy source.
- **capacity factor**: The ratio of actual output over a period of time to the potential output if the plant were to operate at full capacity the entire time.
- **hyperscalers**: Large-scale data center operators that require significant amounts of electricity to support their computing needs.
- **power-purchase agreements**: Long-term contracts between energy buyers and sellers that ensure a stable supply of electricity.
- **carbon-free power**: Electricity generated without emitting carbon dioxide, contributing to sustainability goals.

## Use Cases

- Building data centers with co-located renewable energy sources.
- Utilizing geothermal energy for continuous power supply.
- Signing long-term power-purchase agreements to secure energy.
- Reviving idled nuclear units for energy supply.
- Developing small modular reactors for data center energy needs.
- Implementing behind-the-meter energy generation to bypass grid limitations.

## Frequently Asked Questions

**Why is power considered the bottleneck for AI development?**

Power is the critical constraint because AI models require continuous electricity to operate efficiently. Without reliable and abundant power, the return on investment for expensive computing resources diminishes.

**What role do data centers play in the AI race?**

Data centers are essential for running AI models, as they house the necessary computing infrastructure. Their electricity consumption is projected to grow significantly, making them pivotal in the AI landscape.

**How is geothermal energy relevant to data centers?**

Geothermal energy offers a reliable and carbon-free power source that can be generated close to data centers, addressing both energy demand and sustainability concerns.

**What are the implications of the increasing electricity demand from data centers?**

As data centers demand more electricity, they are becoming aggressive buyers in the energy market, which may lead to innovations in energy procurement and generation strategies to ensure reliability.

**What is the significance of the capacity factor in energy sourcing?**

The capacity factor indicates how reliably an energy source can produce power. For data centers, a high capacity factor is crucial to ensure that GPUs are continuously supplied with electricity, maximizing their efficiency.

## Links

- [Read on Welcome.AI](https://welcome.ai/content/geothermal-energys-role-in-meeting-data-centers-future-power-needs)
- [Original source](https://www.datacenterdynamics.com/en/opinions/the-electrical-power-problem-that-will-decide-the-ai-race/)

---

Source: Welcome.AI | https://welcome.ai/content/geothermal-energys-role-in-meeting-data-centers-future-power-needs