| 1. | EXECUTIVE SUMMARY |
| 1.1. | Scope of IDTechEx's Sustainability for Data Centers Report |
| 1.2. | Data Center CO2 emissions |
| 1.3. | Biggest contribution to data center carbon footprint |
| 1.4. | Summary of data center companies (edge, colocation, hyperscale) |
| 1.5. | Data center hyperscalers: Already biggest buyers of RECs and carbon credits |
| 1.6. | Data center power crisis: Increasingly harder to secure grid connectivity |
| 1.7. | Reality check: Still mostly fossil fuels powering data centers |
| 1.8. | Power Usage Effectiveness (PUE) |
| 1.9. | Data center decarbonization regulation |
| 1.10. | Regional Actions on Reducing GHG Emission |
| 1.11. | Motivations behind data center sustainability and carbon reductions |
| 1.12. | Which sustainable technologies can have the biggest impact? - Low carbon electricity |
| 1.13. | Which sustainable technologies can have the biggest impact? - Efficiency |
| 1.14. | Which sustainable technologies can have the biggest impact? - Scope 3 emissions |
| 1.15. | Benchmarking electricity sources for data centers |
| 1.16. | Wind, solar, and hydropower are all well-established |
| 1.17. | Three applications for Li-ion batteries in data centers |
| 1.18. | Commercial and industrial battery storage technology benchmarking, advantages and disadvantages |
| 1.19. | TDP (thermal design power) trend of GPUs |
| 1.20. | Cooling Technology Comparison |
| 1.21. | Different Cooling on Rack Level |
| 1.22. | AI is driving the push for increased power density, rather than just efficiency |
| 1.23. | Key changes in the 800VDC data center architecture |
| 1.24. | AI chip technologies benchmarked |
| 1.25. | The approach of data center hyperscalers towards CO2 offsetting |
| 1.26. | Scale and technology readiness level of carbon dioxide removal methods |
| 1.27. | EACs (environmental attribute certificates) for low-carbon construction |
| 1.28. | Data Center Power and Electricity Forecast: 2025 to 2036 |
| 1.29. | Data Center CO2 forecast: 2025 to 2036 |
| 1.30. | Access More With an IDTechEx Subscription |
| 2. | INTRODUCTION |
| 2.1. | What is a data center? |
| 2.2. | Different types of data centers |
| 2.3. | Summary of colocation and hyperscaler providers, and data center providers |
| 2.4. | The USA leads significantly on data centers, followed by Germany and the UK |
| 2.5. | Map of global data center hubs |
| 2.6. | Data center sustainability metrics |
| 2.7. | Scope of IDTechEx's Sustainability for Data Centers Report |
| 2.8. | Data center CO2 emissions |
| 2.9. | Motivations behind data center sustainability and carbon reductions |
| 2.10. | Leading global data center hyperscalers |
| 2.11. | Introduction to basic definitions |
| 2.12. | CO2 emissions of data center hyperscalers |
| 2.13. | CO2 emissions of data center hyperscalers - Scope 2: Market-based vs location-based |
| 2.14. | CO2 emissions of data center hyperscalers - Scope 3 |
| 2.15. | GHG emission targets from hyperscalers and colocators |
| 2.16. | Policies for data center decarbonization |
| 2.17. | Power Usage Effectiveness (PUE) |
| 2.18. | Carbon Usage Effectiveness (CUE) |
| 2.19. | Regional Actions on Reducing GHG Emission |
| 2.20. | Hyperscale data centers are the most efficient |
| 2.21. | Low-carbon electricity remains key reduction area for data centers |
| 3. | SUSTAINABLE POWER GENERATION FOR DATA CENTERS |
| 3.1. | Introduction |
| 3.1.1. | Data centers consume large & increasing amounts of power globally |
| 3.1.2. | Data centers need a continuous supply of power |
| 3.1.3. | Data centers get power from the grid |
| 3.1.4. | Carbon intensity of power production varies geographically |
| 3.1.5. | Ways for data centers to decarbonize power and lower Scope 2 emissions |
| 3.1.6. | RECs, PPAs, and Clean Transition Tariffs |
| 3.1.7. | Hyperscale data centers are leading buyers of clean power |
| 3.1.8. | Hourly and location-based power matching |
| 3.1.9. | Comparison of different power sources for data centers |
| 3.1.10. | Benchmarking electricity sources for data centers |
| 3.1.11. | The US Energy Sector: Transitioning from abundance to scarcity |
| 3.1.12. | A significant "power gap" between supply and demand of power for data centers is expected in the US within the next 5 years |
| 3.1.13. | Data center hyperscalers - becoming power generation companies |
| 3.1.14. | Reality check - still mostly fossil fuels powering data centers |
| 3.1.15. | Microgrids and behind-the-meter power for data centers |
| 3.1.16. | Low-carbon microgrids for data centers: Case studies |
| 3.2. | Solar, wind, and hydropower for data centers |
| 3.2.1. | Wind, solar, and hydropower are all well-established |
| 3.2.2. | Challenges for wind and solar - intermittency and footprint |
| 3.2.3. | Space-based solar power and space-based data centers |
| 3.2.4. | Wind and solar for floating off-shore data centers |
| 3.2.5. | Wind power for AI inferencing |
| 3.2.6. | Wind + solar = success? |
| 3.2.7. | Hydropower for data centers |
| 3.3. | Nuclear (conventional, SMRs, and fusion) for data centers |
| 3.3.1. | Nuclear energy has very low carbon emissions compared even to other renewables |
| 3.3.2. | Small modular reactors (SMRs): What and why? |
| 3.3.3. | Nuclear is becoming part of the long-term solution for sustainable data center power, but will not be a near-term quick fix |
| 3.3.4. | SMRs are expected to reduce the cost of nuclear energy |
| 3.3.5. | SMR reactor technologies can be split into Gen III+ and Gen IV designs |
| 3.3.6. | Comparing promising technologies |
| 3.3.7. | Hyperscalers each bet on a different Gen IV SMR startup - except Microsoft |
| 3.3.8. | Data centers are getting bigger, and so are SMR projects |
| 3.3.9. | Opportunities to extend or restart large-scale nuclear power plants |
| 3.3.10. | Building SMRs vs extending/restarting the lifetime of large NPPs |
| 3.3.11. | What is holding back SMRs? |
| 3.3.12. | Analyst opinion: SMR market overview |
| 3.3.13. | Fusion energy: Potential competitor for data center power? |
| 3.3.14. | Fusion power rollout will have a longer time to market than SMRs |
| 3.4. | Geothermal for data centers |
| 3.4.1. | Introduction to geothermal power |
| 3.4.2. | Regions with high geothermal potential |
| 3.4.3. | Introduction to enhanced geothermal systems |
| 3.4.4. | General progress towards commercial EGS |
| 3.4.5. | Economics of enhanced geothermal systems |
| 3.4.6. | Geothermal power for data centers |
| 3.4.7. | Enhanced geothermal systems for data centers: Sage Geosystems and Fervo Energy |
| 3.4.8. | Outlook: Geothermal energy for data centers |
| 3.5. | CCUS (carbon capture, utilization, and storage) for data centers |
| 3.5.1. | Introduction: Carbon capture for data centers |
| 3.5.2. | What is Carbon Capture, Utilization and Storage (CCUS)? |
| 3.5.3. | The momentum behind CCUS is building up |
| 3.5.4. | CCUS milestones in 2025/2026 |
| 3.5.5. | High-concentration CO2 sources are the low-hanging fruits |
| 3.5.6. | CO2 source: From which sectors has CO2 been captured historically? |
| 3.5.7. | Which carbon capture technologies are most mature? |
| 3.5.8. | Point-source carbon capture technology providers |
| 3.5.9. | Power plants with CCUS generate less energy |
| 3.5.10. | CO2 capture from coal power generation |
| 3.5.11. | CO2 capture from gas power generation |
| 3.5.12. | Carbon capture and gas power |
| 3.5.13. | Gas power CCS for data centers |
| 3.6. | Fuel cells for data centers |
| 3.6.1. | Fuel cells for data centers |
| 3.6.2. | Advantages of hydrogen |
| 3.6.3. | What are fuel cells? |
| 3.6.4. | Proton exchange membrane fuel cell (PEMFC) technology overview |
| 3.6.5. | Solid oxide fuel cell (SOFC) technology overview |
| 3.6.6. | Data centers: PEMFCs and SOFCs |
| 3.6.7. | Data centres and telecom application technology considerations |
| 3.6.8. | Technology benchmarking for data centres applications |
| 3.6.9. | The hydrogen economy and its impact on the fuel cell market |
| 3.6.10. | Status of hydrogen |
| 3.6.11. | Barriers remain for low-carbon hydrogen |
| 3.6.12. | Key players for PEMFC and SOFC stationary fuel cells |
| 3.6.13. | PEMFC data center case studies: Projects use hydrogen |
| 3.6.14. | SOFC data center case studies: Projects use natural gas instead of hydrogen |
| 3.6.15. | Outlook: Hydrogen fuel cells in data centers will remain niche |
| 3.7. | Batteries and energy storage for data centers |
| 3.7.1. | Three sustainability applications for Li-ion batteries in data centers |
| 3.7.2. | UPS battery technologies |
| 3.7.3. | Li-ion UPS case study: Ampace |
| 3.7.4. | Grid-interactive UPS technologies |
| 3.7.5. | Batteries for back-up power generation in data centers: Case studies |
| 3.7.6. | Battery renewable energy storage for data center primary power |
| 3.7.7. | Key players in grid-scale Li-ion BESS |
| 3.7.8. | Microgrids and batteries for data center primary power |
| 3.7.9. | Commercial and industrial battery storage technology benchmarking, advantages and disadvantages |
| 3.7.10. | Redox flow batteries for data centers |
| 3.7.11. | Key battery storage data center market developments summary |
| 3.7.12. | Emerging LDES (long-duration energy storage) for data centers |
| 3.7.13. | TRL of LDES solutions |
| 3.7.14. | CO2 Battery: Carbon dioxide utilization for energy storage |
| 4. | ENERGY EFFICIENCY FOR DATA CENTERS |
| 4.1. | Introduction |
| 4.1.1. | Chapter overview |
| 4.1.2. | Avenues to improve efficiencies |
| 4.1.3. | "If PUE was the answer, we wouldn't still be talking about data center sustainability" |
| 4.2. | Thermal efficiency (thermal management for data centers) |
| 4.2.1. | Thermal management trends and impact on data center sustainability |
| 4.2.2. | TDP (thermal design power) trend of GPUs |
| 4.2.3. | Overview of cooling in data centers |
| 4.2.4. | Cooling Technology Comparison |
| 4.2.5. | Evolution of cooling technologies for data centers |
| 4.2.6. | Liquid Cooling - Power Limitation of Different Cooling on Rack Level |
| 4.2.7. | Different Cooling on Rack Level |
| 4.2.8. | Data centers using refrigerants with lower GWP |
| 4.2.9. | Introduction to D2C cooling |
| 4.2.10. | Single and two-phase cold plate cooling: Challenges |
| 4.2.11. | Cost analysis of Two-Phase Direct-to-Chip Cooling: Accelsius |
| 4.2.12. | Single-phase and two-phase immersion |
| 4.2.13. | Single and two-phase immersion cooling: Challenges |
| 4.2.14. | Data Center Cooling Value Chain |
| 4.2.15. | Quantifying Efficiency Improvements - Air vs Direct-to-Chip Cooling |
| 4.2.16. | GHG emission by cooling method |
| 4.2.17. | Themoelectric cooling for data centers |
| 4.2.18. | Waste heat reuse: Encouraged now, mandated soon? |
| 4.2.19. | Data centers for district heating: Europe leads |
| 4.2.20. | Maximizing data center efficiency and sustainability - WEDistrict project |
| 4.3. | Power efficiency (power supply, 800VDC) |
| 4.3.1. | Key takeaways from this subchapter |
| 4.3.2. | Some notes on data center architecture |
| 4.3.3. | Power electronics play a critical role in data center operation |
| 4.3.4. | Power supply units are critical for data center efficiency and power density |
| 4.3.5. | Power factor correction is a crucial use of power electronics |
| 4.3.6. | The "80 Plus" program sets the global standard for PSU efficiency |
| 4.3.7. | Other programs incentivize or mandate efficiency requirements globally |
| 4.3.8. | Adoption of SiC and GaN in the PSU |
| 4.3.9. | Comparison of semiconductor material properties |
| 4.3.10. | Comparison of Si to wide bandgap semiconductors SiC and GaN |
| 4.3.11. | Timeline of SiC and GaN adoption in PSUs |
| 4.3.12. | AI is driving the push for increased power density, rather than just efficiency |
| 4.3.13. | Rack densities in data centers will continue to increase |
| 4.3.14. | Existing data center architecture cannot support future server generations |
| 4.3.15. | Key changes in the 800VDC data center architecture |
| 4.3.16. | NVIDIA partners with numerous WBG Tier 1s for new data centers |
| 4.3.17. | Summary of next-generation 8-12kW PSUs |
| 4.3.18. | 800VDC rack architecture: Retrofit vs new |
| 4.3.19. | Energy efficiency of traditional AC vs 800VDC architecture |
| 4.3.20. | The concept of HTS for power distribution in data centers has emerged recently |
| 4.3.21. | Technical primer: High-temperature superconductors (HTS tapes) |
| 4.3.22. | HTS tapes could replace copper busbars in data centers |
| 4.3.23. | Structure of a HTS transmission line |
| 4.3.24. | Performance improvements from using HTS for power distribution |
| 4.3.25. | Advantages of deploying HTS for power distribution in data centers |
| 4.3.26. | Challenges of deploying HTS for power distribution in data centers |
| 4.4. | IT efficiency (AI chips and storage) |
| 4.4.1. | Key takeaways from this subchapter |
| 4.4.2. | AI chip technologies benchmarked |
| 4.4.3. | Technology readiness of AI chip technologies |
| 4.4.4. | AI chip landscape - Chip designers |
| 4.4.5. | Performance per watt for different AI chips |
| 4.4.6. | Future outlook - the search for flexible architectures with high energy efficiency and performance |
| 4.4.7. | Electricity (operational emissions) is the biggest contributor to the carbon footprint of AI chips |
| 4.4.8. | Increasing SSD capacity through emerging lower cost QLC NAND |
| 4.4.9. | Energy efficiency of SSD storage technologies for datacenter and enterprise |
| 4.4.10. | TRL of storage technologies for HPC Datacenter and Enterprise |
| 5. | ADDITIONAL SCOPE 3 DECARBONIZATION FOR DATA CENTERS |
| 5.1. | Introduction |
| 5.1.1. | CO2 emissions of data centers: Scope 3 emissions |
| 5.2. | Carbon credits/CO2 offsetting |
| 5.2.1. | What is a carbon credit and carbon offsetting? |
| 5.2.2. | Carbon removal vs carbon avoidance offsetting |
| 5.2.3. | The approach of data center hyperscalers towards CO2 offsetting |
| 5.2.4. | Overall picture: Voluntary carbon credit markets |
| 5.2.5. | The carbon removal supply chain: Carbon credit market players |
| 5.2.6. | How are voluntary carbon credits purchased? |
| 5.2.7. | Data center hyperscalers are the biggest durable carbon removal buyers |
| 5.2.8. | How much do durable carbon removal credits in cost? |
| 5.2.9. | Scale and technology readiness level of carbon dioxide removal methods |
| 5.2.10. | Pre-purchases dominate the durable CDR space |
| 5.2.11. | Carbon dioxide removal market trends in 2025: Nature-based removals |
| 5.2.12. | Carbon dioxide removal market trends in 2025: Durable, engineered removals |
| 5.2.13. | Increasing buyer preference for durable CDR |
| 5.2.14. | CDR and compliance carbon markets |
| 5.2.15. | CDR portfolios of data center hyperscalers |
| 5.2.16. | What are the major challenges for scaling up direct air capture? |
| 5.2.17. | DAC technology landscape: Companies |
| 5.2.18. | Which DAC technologies will be the most successful? |
| 5.2.19. | Most existing BECCS projects are in ethanol production |
| 5.2.20. | Solvent capture technologies dominate the BECCS space |
| 5.2.21. | The potential for BiCRS goes beyond BECCS: Benchmarking |
| 5.2.22. | The state of the global biochar market |
| 5.2.23. | Biochar CDR is scaling up |
| 5.3. | Green concrete and green steel: Low-carbon construction |
| 5.3.1. | Construction of data centers is increasing |
| 5.3.2. | EACs (environmental attribute certificates) for low-carbon construction |
| 5.3.3. | Contribution of concrete to data center CO2 footprint |
| 5.3.4. | How much does data center construction cost? |
| 5.3.5. | Precast concrete in data center construction |
| 5.3.6. | Low-carbon concrete requires cement decarbonization technologies |
| 5.3.7. | Benchmarking cement decarbonization technologies |
| 5.3.8. | How much will the green premium be for decarbonized cement? |
| 5.3.9. | Cement decarbonization - Analyst viewpoint: Benchmarking of cement decarbonization technologies |
| 5.3.10. | Data center hyperscalers have formed partnerships with low-carbon concrete players |
| 5.3.11. | Role of steel in data centers |
| 5.3.12. | Overview of decarbonization technologies for the steel sector |
| 5.3.13. | Tech companies' interest in green steel |
| 5.3.14. | Data center hyperscalers have formed partnerships with low-carbon steel players |
| 5.3.15. | Cross-laminated timber as a low-carbon construction material |
| 5.4. | IT and embodied carbon |
| 5.4.1. | Scope 3 data center emissions - the importance of embodied carbon |
| 5.4.2. | IT makes the largest contribution to embodied carbon |
| 5.4.3. | Manufacturing/Embodied GHG Emission - Componentry Level Split of Servers |
| 5.4.4. | Manufacturing/Embodied GHG Emission - GPU baseboard |
| 5.4.5. | IT hardware reuse in data centers |
| 6. | MARKET FORECASTS |
| 6.1. | Global data center forecasts |
| 6.1.1. | Data Center Power and Electricity Forecast: 2023 to 2036 |
| 6.1.2. | Data Center CO2 forecast: 2025 to 2036 |
| 6.1.3. | TDP Trend: Historic Data and Forecast Data - GPU |
| 6.1.4. | Revenue forecast by cooling method |
| 6.1.5. | Power forecast 800VDC (HVDC) |
| 6.2. | Global green market forecasts |
| 6.2.1. | Low-carbon steel forecast by technology, 2025-2035 |
| 6.2.2. | Carbon dioxide removal capacity forecast by technology (million metric tons of CO2 per year), 2025-2035 |
| 6.2.3. | Nuclear SMR forecast by technology, 2026-2027 |
| 6.2.4. | Forecasting methodology |
| 7. | COMPANY PROFILES |
| 7.1. | AcBel Polytech Inc. |
| 7.2. | Ampace: UPS |
| 7.3. | Atecom Technology Co., Ltd. |
| 7.4. | Ballard Power Systems |
| 7.5. | Boston Metal |
| 7.6. | CATL: Battery Energy Storage Systems (BESS) |
| 7.7. | Ceres |
| 7.8. | Climeworks |
| 7.9. | Electra |
| 7.10. | FlexBase |
| 7.11. | Fluence: Battery Energy Storage Systems (BESS) |
| 7.12. | Hebei Synlight Semiconductor |
| 7.13. | Intelligent Energy |
| 7.14. | Kairos Power |
| 7.15. | Mission Zero Technologies |
| 7.16. | Mitsubishi Heavy Industries: KM CDR Process |
| 7.17. | Narada Power |
| 7.18. | Nedstack |
| 7.19. | Oklo |
| 7.20. | Orien Energy |
| 7.21. | Phlair |
| 7.22. | Plug Power |
| 7.23. | PowerCell |
| 7.24. | SLB Capturi |
| 7.25. | Stegra (H2 Green Steel) |
| 7.26. | Sublime Systems |
| 7.27. | Sungrow |
| 7.28. | TerraPower Natrium |
| 7.29. | TYSiC |
| 7.30. | Vertiv Holdings — Data Center Liquid Cooling |
| 7.31. | Vision Electric Superconductors |
| 7.32. | Wärtsilä |
| 7.33. | X-Energy |