Back to overview

Back to overview

Google Reroutes Compute to Chase Clean Energy Around the World

Google's carbon-aware computing platform uses Electricity Maps' hourly carbon intensity forecasts to smartly shift computing tasks

Jun 18, 2025

Jesper Klingenberg

Tackling the climate challenge through smarter computing

‍The fight against climate change requires innovation across every layer of technology. For Google, that means ensuring that even the invisible parts of the internet — the servers powering everything from search to YouTube to Google Cloud — operate in ways that minimize environmental impact.

Google has committed to running its data centers and campuses on 24/7 carbon-free energy (CFE) — meaning they aim to match their electricity consumption with clean energy, locally and hourly, every hour of every day. This goes far beyond annual offsets or credits; it’s about real-time alignment between electricity use and low-carbon generation.

The role of Electricity Maps

Electricity Maps provides hourly carbon intensity forecasts for grids around the world — using a combination of live grid data, weather patterns, and energy modeling. This data gives Google precise visibility into when and where electricity is cleanest — whether it’s driven by wind in Denmark, hydro in Oregon, or solar in Chile.

This real-time and forecasted data enables Google’s carbon-intelligent computing platform to make smart decisions on when to execute flexible compute tasks, like data analytics, training AI models, or batch processing.

From data to action: shifting workloads intelligently

To act on this data, Google’s platform compares two inputs:

  • Carbon intensity forecasts from Electricity Maps — predicting how clean or dirty the local grid will be in the coming hours.

  • Compute load forecasts from internal systems — estimating how much energy each data center will need across the same period.

With this information, Google can generate hourly guidelines for when to prioritize power-intensive tasks. These decisions are then passed on to workload schedulers, allowing jobs to be automatically moved to times or locations with cleaner energy — all while maintaining performance and uptime.

In some cases, if the local grid is forecasted to be high in emissions, workloads can be shifted geographically — from one data center to another, in a region with greener energy. This global orchestration lets Google optimize its compute operations not just in time, but also in space.

Real-world impact

This system has already been deployed in several of Google’s largest data centers and is being scaled globally. The impact is meaningful:


  • Lower CO₂ emissions per compute cycle

  • Better use of existing clean energy without needing to build new infrastructure

  • More resilient systems that can respond to dynamic changes in the grid

As the grid continues to decarbonize and variability from renewables increases, carbon-aware computing becomes an essential tool — enabling Google to maximize its use of carbon-free energy while setting an industry example for sustainable operations.

Tackling the climate challenge through smarter computing

‍The fight against climate change requires innovation across every layer of technology. For Google, that means ensuring that even the invisible parts of the internet — the servers powering everything from search to YouTube to Google Cloud — operate in ways that minimize environmental impact.

Google has committed to running its data centers and campuses on 24/7 carbon-free energy (CFE) — meaning they aim to match their electricity consumption with clean energy, locally and hourly, every hour of every day. This goes far beyond annual offsets or credits; it’s about real-time alignment between electricity use and low-carbon generation.

The role of Electricity Maps

Electricity Maps provides hourly carbon intensity forecasts for grids around the world — using a combination of live grid data, weather patterns, and energy modeling. This data gives Google precise visibility into when and where electricity is cleanest — whether it’s driven by wind in Denmark, hydro in Oregon, or solar in Chile.

This real-time and forecasted data enables Google’s carbon-intelligent computing platform to make smart decisions on when to execute flexible compute tasks, like data analytics, training AI models, or batch processing.

From data to action: shifting workloads intelligently

To act on this data, Google’s platform compares two inputs:

  • Carbon intensity forecasts from Electricity Maps — predicting how clean or dirty the local grid will be in the coming hours.

  • Compute load forecasts from internal systems — estimating how much energy each data center will need across the same period.

With this information, Google can generate hourly guidelines for when to prioritize power-intensive tasks. These decisions are then passed on to workload schedulers, allowing jobs to be automatically moved to times or locations with cleaner energy — all while maintaining performance and uptime.

In some cases, if the local grid is forecasted to be high in emissions, workloads can be shifted geographically — from one data center to another, in a region with greener energy. This global orchestration lets Google optimize its compute operations not just in time, but also in space.

Real-world impact

This system has already been deployed in several of Google’s largest data centers and is being scaled globally. The impact is meaningful:


  • Lower CO₂ emissions per compute cycle

  • Better use of existing clean energy without needing to build new infrastructure

  • More resilient systems that can respond to dynamic changes in the grid

As the grid continues to decarbonize and variability from renewables increases, carbon-aware computing becomes an essential tool — enabling Google to maximize its use of carbon-free energy while setting an industry example for sustainable operations.

Tackling the climate challenge through smarter computing

‍The fight against climate change requires innovation across every layer of technology. For Google, that means ensuring that even the invisible parts of the internet — the servers powering everything from search to YouTube to Google Cloud — operate in ways that minimize environmental impact.

Google has committed to running its data centers and campuses on 24/7 carbon-free energy (CFE) — meaning they aim to match their electricity consumption with clean energy, locally and hourly, every hour of every day. This goes far beyond annual offsets or credits; it’s about real-time alignment between electricity use and low-carbon generation.

The role of Electricity Maps

Electricity Maps provides hourly carbon intensity forecasts for grids around the world — using a combination of live grid data, weather patterns, and energy modeling. This data gives Google precise visibility into when and where electricity is cleanest — whether it’s driven by wind in Denmark, hydro in Oregon, or solar in Chile.

This real-time and forecasted data enables Google’s carbon-intelligent computing platform to make smart decisions on when to execute flexible compute tasks, like data analytics, training AI models, or batch processing.

From data to action: shifting workloads intelligently

To act on this data, Google’s platform compares two inputs:

  • Carbon intensity forecasts from Electricity Maps — predicting how clean or dirty the local grid will be in the coming hours.

  • Compute load forecasts from internal systems — estimating how much energy each data center will need across the same period.

With this information, Google can generate hourly guidelines for when to prioritize power-intensive tasks. These decisions are then passed on to workload schedulers, allowing jobs to be automatically moved to times or locations with cleaner energy — all while maintaining performance and uptime.

In some cases, if the local grid is forecasted to be high in emissions, workloads can be shifted geographically — from one data center to another, in a region with greener energy. This global orchestration lets Google optimize its compute operations not just in time, but also in space.

Real-world impact

This system has already been deployed in several of Google’s largest data centers and is being scaled globally. The impact is meaningful:


  • Lower CO₂ emissions per compute cycle

  • Better use of existing clean energy without needing to build new infrastructure

  • More resilient systems that can respond to dynamic changes in the grid

As the grid continues to decarbonize and variability from renewables increases, carbon-aware computing becomes an essential tool — enabling Google to maximize its use of carbon-free energy while setting an industry example for sustainable operations.

Tackling the climate challenge through smarter computing

‍The fight against climate change requires innovation across every layer of technology. For Google, that means ensuring that even the invisible parts of the internet — the servers powering everything from search to YouTube to Google Cloud — operate in ways that minimize environmental impact.

Google has committed to running its data centers and campuses on 24/7 carbon-free energy (CFE) — meaning they aim to match their electricity consumption with clean energy, locally and hourly, every hour of every day. This goes far beyond annual offsets or credits; it’s about real-time alignment between electricity use and low-carbon generation.

The role of Electricity Maps

Electricity Maps provides hourly carbon intensity forecasts for grids around the world — using a combination of live grid data, weather patterns, and energy modeling. This data gives Google precise visibility into when and where electricity is cleanest — whether it’s driven by wind in Denmark, hydro in Oregon, or solar in Chile.

This real-time and forecasted data enables Google’s carbon-intelligent computing platform to make smart decisions on when to execute flexible compute tasks, like data analytics, training AI models, or batch processing.

From data to action: shifting workloads intelligently

To act on this data, Google’s platform compares two inputs:

  • Carbon intensity forecasts from Electricity Maps — predicting how clean or dirty the local grid will be in the coming hours.

  • Compute load forecasts from internal systems — estimating how much energy each data center will need across the same period.

With this information, Google can generate hourly guidelines for when to prioritize power-intensive tasks. These decisions are then passed on to workload schedulers, allowing jobs to be automatically moved to times or locations with cleaner energy — all while maintaining performance and uptime.

In some cases, if the local grid is forecasted to be high in emissions, workloads can be shifted geographically — from one data center to another, in a region with greener energy. This global orchestration lets Google optimize its compute operations not just in time, but also in space.

Real-world impact

This system has already been deployed in several of Google’s largest data centers and is being scaled globally. The impact is meaningful:


  • Lower CO₂ emissions per compute cycle

  • Better use of existing clean energy without needing to build new infrastructure

  • More resilient systems that can respond to dynamic changes in the grid

As the grid continues to decarbonize and variability from renewables increases, carbon-aware computing becomes an essential tool — enabling Google to maximize its use of carbon-free energy while setting an industry example for sustainable operations.

“Electricity Maps has been a valuable partner, and their expertise in grid carbon data has provided vital support to our clean energy efforts. We have already learned and achieved a great deal through our collaboration, and we look forward to pursuing more great projects together.”

Savannah Goodman
Data and Software Climate Solutions Lead

“Electricity Maps has been a valuable partner, and their expertise in grid carbon data has provided vital support to our clean energy efforts. We have already learned and achieved a great deal through our collaboration, and we look forward to pursuing more great projects together.”

Savannah Goodman
Data and Software Climate Solutions Lead

“Electricity Maps has been a valuable partner, and their expertise in grid carbon data has provided vital support to our clean energy efforts. We have already learned and achieved a great deal through our collaboration, and we look forward to pursuing more great projects together.”

Savannah Goodman
Data and Software Climate Solutions Lead

“Electricity Maps has been a valuable partner, and their expertise in grid carbon data has provided vital support to our clean energy efforts. We have already learned and achieved a great deal through our collaboration, and we look forward to pursuing more great projects together.”

Savannah Goodman
Data and Software Climate Solutions Lead

The collaboration between Google and Electricity Maps is a powerful example of how real-time data and intelligent systems can turn climate ambition into tangible progress. By embedding carbon awareness into the very fabric of its infrastructure, Google is not only reducing its own emissions footprint — it’s actively demonstrating how digital infrastructure can adapt to the realities of a decarbonizing grid. Electricity Maps’ granular and localized carbon intensity data empowers Google to make precise, hour-by-hour decisions about where and when to run compute workloads, aligning demand with cleaner supply across both time and geography.

The collaboration between Google and Electricity Maps is a powerful example of how real-time data and intelligent systems can turn climate ambition into tangible progress. By embedding carbon awareness into the very fabric of its infrastructure, Google is not only reducing its own emissions footprint — it’s actively demonstrating how digital infrastructure can adapt to the realities of a decarbonizing grid. Electricity Maps’ granular and localized carbon intensity data empowers Google to make precise, hour-by-hour decisions about where and when to run compute workloads, aligning demand with cleaner supply across both time and geography.

The collaboration between Google and Electricity Maps is a powerful example of how real-time data and intelligent systems can turn climate ambition into tangible progress. By embedding carbon awareness into the very fabric of its infrastructure, Google is not only reducing its own emissions footprint — it’s actively demonstrating how digital infrastructure can adapt to the realities of a decarbonizing grid. Electricity Maps’ granular and localized carbon intensity data empowers Google to make precise, hour-by-hour decisions about where and when to run compute workloads, aligning demand with cleaner supply across both time and geography.

The collaboration between Google and Electricity Maps is a powerful example of how real-time data and intelligent systems can turn climate ambition into tangible progress. By embedding carbon awareness into the very fabric of its infrastructure, Google is not only reducing its own emissions footprint — it’s actively demonstrating how digital infrastructure can adapt to the realities of a decarbonizing grid. Electricity Maps’ granular and localized carbon intensity data empowers Google to make precise, hour-by-hour decisions about where and when to run compute workloads, aligning demand with cleaner supply across both time and geography.

What makes this approach so impactful is that it doesn’t rely on future technologies or hypothetical offsets — it’s operational today, running at scale across some of the most complex and energy-intensive computing environments in the world. The system’s ability to shift non-urgent compute tasks without compromising service reliability opens up a powerful lever for decarbonization that other industries can learn from. In a world where clean energy is increasingly abundant but often variable, strategies like carbon-aware computing ensure we make the most of every green electron already available.

This partnership points toward a new operational paradigm where energy and compute scheduling are no longer disconnected — and where sustainability becomes a built-in feature rather than an external goal. As more organizations pursue 24/7 carbon-free energy, the collaboration between Google and Electricity Maps offers a working model for what it takes: rich, real-time data; intelligent automation; and a willingness to reimagine long-held assumptions about infrastructure and energy use. What started as a technical integration is now shaping the future of sustainable digital operations — and it’s only the beginning.

What makes this approach so impactful is that it doesn’t rely on future technologies or hypothetical offsets — it’s operational today, running at scale across some of the most complex and energy-intensive computing environments in the world. The system’s ability to shift non-urgent compute tasks without compromising service reliability opens up a powerful lever for decarbonization that other industries can learn from. In a world where clean energy is increasingly abundant but often variable, strategies like carbon-aware computing ensure we make the most of every green electron already available.

This partnership points toward a new operational paradigm where energy and compute scheduling are no longer disconnected — and where sustainability becomes a built-in feature rather than an external goal. As more organizations pursue 24/7 carbon-free energy, the collaboration between Google and Electricity Maps offers a working model for what it takes: rich, real-time data; intelligent automation; and a willingness to reimagine long-held assumptions about infrastructure and energy use. What started as a technical integration is now shaping the future of sustainable digital operations — and it’s only the beginning.

What makes this approach so impactful is that it doesn’t rely on future technologies or hypothetical offsets — it’s operational today, running at scale across some of the most complex and energy-intensive computing environments in the world. The system’s ability to shift non-urgent compute tasks without compromising service reliability opens up a powerful lever for decarbonization that other industries can learn from. In a world where clean energy is increasingly abundant but often variable, strategies like carbon-aware computing ensure we make the most of every green electron already available.

This partnership points toward a new operational paradigm where energy and compute scheduling are no longer disconnected — and where sustainability becomes a built-in feature rather than an external goal. As more organizations pursue 24/7 carbon-free energy, the collaboration between Google and Electricity Maps offers a working model for what it takes: rich, real-time data; intelligent automation; and a willingness to reimagine long-held assumptions about infrastructure and energy use. What started as a technical integration is now shaping the future of sustainable digital operations — and it’s only the beginning.

What makes this approach so impactful is that it doesn’t rely on future technologies or hypothetical offsets — it’s operational today, running at scale across some of the most complex and energy-intensive computing environments in the world. The system’s ability to shift non-urgent compute tasks without compromising service reliability opens up a powerful lever for decarbonization that other industries can learn from. In a world where clean energy is increasingly abundant but often variable, strategies like carbon-aware computing ensure we make the most of every green electron already available.

This partnership points toward a new operational paradigm where energy and compute scheduling are no longer disconnected — and where sustainability becomes a built-in feature rather than an external goal. As more organizations pursue 24/7 carbon-free energy, the collaboration between Google and Electricity Maps offers a working model for what it takes: rich, real-time data; intelligent automation; and a willingness to reimagine long-held assumptions about infrastructure and energy use. What started as a technical integration is now shaping the future of sustainable digital operations — and it’s only the beginning.

Related articles

Continue reading

Electricity Maps logo

Back

Pricing

Technology

Careers

6

Features

Coverage

Signals

Temporality

Products

Interactive Map

Datasets

Developer Portal

Company

About

Team

News & events

Contact us

Sign in

Sign up

Electricity Maps logo

Back

Pricing

Technology

Careers

6

Features

Coverage

Signals

Temporality

Products

Interactive Map

Datasets

Developer Portal

Company

About

Team

News & events

Contact us

Sign in

Sign up