Growing the internet while reducing energy consumption
We strive to build the world’s most energy-efficient computing infrastructure, outfitting each Google data center with high-performance servers designed to use as little energy as possible compared to the amount of data they process.
Google Data Center PUE measurement boundaries. In 2025, the average annual power usage effectiveness for our global fleet of data centers was 1.09.1
Google’s comprehensive approach to measuring Power Usage Effectiveness (PUE)
Our calculations are based on continuously measuring the entire worldwide fleet performance of our data centers throughout the year. We report a comprehensive trailing twelve-month (TTM) PUE of 1.09 across all our large-scale data centers (once they reach stable operations), in all seasons, including all sources of overhead.
Google’s data centers are among the industry’s most efficient, operating with a 2025 fleet-wide average PUE of 1.09 and using 83% less overhead energy than the industry average1. In 2025, our data centers delivered over 3 times more compute performance per unit of energy than five years ago, driven largely by TPU energy efficiency improvements and their deployment to support AI.2
1PUE (power usage effectiveness) is a standard industry ratio that compares the amount of non-computing overhead energy (used for things like cooling and power distribution) to the amount of energy used to powerIT equipment. For example, a PUE of 2.0 means that for every watt of IT power, an additional watt is consumed to cool and distribute power to the IT equipment. A PUE closer to 1.0 means nearly all the energy is used for computing. According to the Uptime Institute’s 2025 Global Data Center Survey, the global average PUE of respondents’ data centers was 1.54. The overhead energy use comparison was calculated as follows: (1 - (Google’s overhead energy use [0.09] divided by the industry average overhead energy use [0.54])) x 100 = 83%.
2We’ve updated our methodology for estimating our fleet-wide compute efficiency. This is based on internal analysis of the estimated energy consumption required for comparable work with CPU and GPU/TPU hardware from 2020 compared to 2025.
2026 PUE Yearly Report
| Fleet wide PUE | Quarterly PUE | Trailing twelve-month (TTM) PUE |
| Fleet | 1.08 | 1.09 |
| Campuses | Quarterly PUE | Trailing twelve-month (TTM) PUE |
| Berkeley County, South Carolina | 1.08 | 1.09 |
| Bristow, Virginia | 1.07 | 1.09 |
| Central Ohio (Columbus), Ohio | 1.06 | 1.06 |
| Central Ohio (Lancaster), Ohio | 1.04 | 1.04 |
| Central Ohio (New Albany), Ohio | 1.07 | 1.06 |
| Changua County, Taiwan | 1.11 | 1.13 |
| Council Bluffs, Iowa | 1.1 | 1.11 |
| Council Bluffs, Iowa (2nd facility) | 1.07 | 1.08 |
| Douglas County, Georgia | 1.07 | 1.09 |
| Dublin, Ireland | 1.08 | 1.08 |
| Eemshaven, Netherlands | 1.07 | 1.07 |
| Fredericia, Denmark | 1.06 | 1.06 |
| Hamina, Finland | 1.1 | 1.1 |
| Henderson, Nevada | 1.08 | 1.09 |
| Inzai, Japan | 1.08 | 1.1 |
| Jackson County, Alabama | 1.09 | 1.1 |
| Lenoir, North Carolina | 1.08 | 1.1 |
| Loudoun County, Virginia | 1.07 | 1.08 |
| Loudoun County, Virginia (2nd facility) | 1.08 | 1.08 |
| Mayes County, Oklahoma | 1.08 | 1.11 |
| Mesa, Arizona | 1.22 | |
| Midlothian, Texas | 1.08 | 1.1 |
| Montgomery County, Tennessee | 1.09 | 1.09 |
| Omaha, Nebraska | 1.05 | 1.05 |
| Papillion, Nebraska | 1.07 | 1.09 |
| Quilicura, Chile | 1.08 | 1.08 |
| Red Oak, Texas | 1.1 | |
| Singapore | 1.12 | 1.12 |
| Singapore (2nd facility) | 1.14 | 1.14 |
| St. Ghislain, Belgium | 1.07 | 1.08 |
| Storey County, Nevada | 1.11 | 1.15 |
| The Dalles, Oregon | 1.1 | 1.1 |
| The Dalles, Oregon (2nd facility) | 1.06 | 1.06 |
| The Dalles, Oregon (3rd facility) | 1.07 |
*We report individual campus TTM PUE only for campuses with at least twelve months of data.
| Fleet wide PUE | Quarterly PUE | Trailing twelve-month (TTM) PUE |
| Fleet | 1.10 | 1.09 |
| Campuses | Quarterly PUE | Trailing twelve-month (TTM) PUE |
| Berkeley County, South Carolina | 1.11 | 1.10 |
| Bristow, Virginia | 1.10 | 1.09 |
| Central Ohio (Columbus), Ohio | 1.06 | 1.06 |
| Central Ohio (Lancaster), Ohio | 1.04 | 1.04 |
| Central Ohio (New Albany), Ohio | 1.08 | 1.07 |
| Changua County, Taiwan | 1.13 | 1.13 |
| Council Bluffs, Iowa | 1.12 | 1.11 |
| Council Bluffs, Iowa (2nd facility) | 1.08 | 1.08 |
| Douglas County, Georgia | 1.10 | 1.09 |
| Dublin, Ireland | 1.08 | 1.08 |
| Eemshaven, Netherlands | 1.07 | 1.07 |
| Fort Wayne, Indiana | 1.05 | |
| Fredericia, Denmark | 1.06 | 1.06 |
| Hamina, Finland | 1.09 | 1.09 |
| Henderson, Nevada | 1.10 | 1.09 |
| Inzai, Japan | 1.11 | 1.10 |
| Jackson County, Alabama | 1.12 | 1.11 |
| Lenoir, North Carolina | 1.10 | 1.10 |
| Loudoun County, Virginia | 1.09 | 1.08 |
| Loudoun County, Virginia (2nd facility) | 1.09 | 1.08 |
| Mayes County, Oklahoma | 1.11 | 1.11 |
| Mesa, Arizona | 1.28 | |
| Midlothian, Texas | 1.12 | 1.11 |
| Montgomery County, Tennessee | 1.12 | 1.09 |
| Omaha, Nebraska | 1.06 | 1.05 |
| Papillion, Nebraska | 1.09 | 1.09 |
| Quilicura, Chile | 1.07 | 1.08 |
| Red Oak, Texas | 1.14 | |
| Singapore | 1.13 | 1.12 |
| Singapore (2nd facility) | 1.17 | 1.15 |
| St. Ghislain, Belgium | 1.08 | 1.08 |
| Storey County, Nevada | 1.17 | 1.15 |
| The Dalles, Oregon | 1.10 | 1.10 |
| The Dalles, Oregon (2nd facility) | 1.07 | 1.06 |
| The Dalles, Oregon (3rd facility) | 1.08 | 1.08 |
*We report individual campus TTM PUE only for campuses with at least twelve months of data.
2025 PUE Yearly Report
| Fleet wide PUE | Quarterly PUE | Trailing twelve-month (TTM) PUE |
| Fleet | 1.08 | 1.09 |
| Campuses | Quarterly PUE | Trailing twelve-month (TTM) PUE |
| Berkeley County, South Carolina | 1.07 | 1.1 |
| Bristow, Virginia | 1.07 | |