Prepare and test your skills

Prepare and test your skills

Google Cloud (opens in a new tab) enables sustainable operations at scale by providing carbon-neutral infrastructure, efficient data centers, and measurement tools that help organizations grow while lowering environmental impact. Google manages efficiency of the platform through renewable energy, advanced cooling, and hardware optimization, while customers manage efficiency in the cloud by running lean workloads and tracking emissions. This partnership lets organizations meet environmental commitments without giving up performance, cost control, or scale.
The sustainability value of migrating to Google Cloud is lower energy use and lower associated emissions for the same work. Migrating on-premises workloads to Google Cloud reduces energy use and associated emissions by 1.4 to 2 times compared with typical on-premises deployments. Google Cloud data centers are custom-designed for high power usage effectiveness (PUE), which measures how efficiently a facility uses energy for computing versus cooling and power distribution. Older on-premises rooms often lack the scale to justify advanced cooling and power systems, so they consume more energy for the same computing tasks.
Shared responsibility for sustainable operations at scale divides work at the boundary between Google Cloud infrastructure and customer workloads. Google is responsible for sustainability of Google Cloud, and the customer is responsible for sustainability in the cloud, with shared fate meaning both sides work as partners to reduce the footprint of the whole ecosystem. Customer workloads depend on Google Cloud data centers for efficient power, cooling, and carbon-free energy, while Google depends on customers to avoid idle resources and oversized machines.
| Owner | Responsible for | What it includes |
|---|---|---|
| Sustainability of Google Cloud | Energy efficiency and water stewardship of data centers, infrastructure, and core services, plus renewable energy, climate-conscious cooling, and hardware optimization | |
| Customer | Sustainability in the cloud | Energy efficiency of workloads, including right-sizing resources, using serverless services that scale to zero, and managing data lifecycles |
Exam tip: Sustainability of Google Cloud means Google's data centers and core services, while sustainability in the cloud means the customer's workloads and resource choices.
Carbon Footprint (opens in a new tab) is the Google Cloud tool that displays gross greenhouse gas emissions from electricity associated with covered Google Cloud services. Emissions data flows from Carbon Footprint to the console dashboard for review by project, region, and service, and to BigQuery or Sheets for deeper analysis, custom dashboards, and reports. The report uses two accounting methods together so teams see both physical impact and clean-energy benefit. Location-based accounting reflects the emissions factor of the local electrical grid, while market-based accounting reflects the benefit of Google's carbon-free energy purchases.
This dual view helps organizations choose low-carbon regions, prioritize high-emission projects, and verify that actions such as right-sizing or decommissioning actually lower emissions. Sustainable-by-design choices also reinforce other goals, because faster performance uses compute for less time, lower cost means fewer resources consumed, strong security prevents energy-intensive incidents, and efficient applications reduce energy on network and end-user devices.
Data and AI products that support sustainable operations turn large planetary, weather, and location datasets into actions for planning, risk management, and efficient operations. Data flows from Earth, weather, and mapping sources to decision-makers who use the insights for site selection, logistics, and climate resilience. These products move organizations from reactive compliance to proactive optimization at scale:
A real-world example is the collaboration between Google Cloud and TELUS, where workloads were migrated to carbon-neutral infrastructure and data analytics were used to optimize operations. That work supports smart-city technology that uses real-time data to reduce traffic congestion and carbon emissions across municipalities in Canada.
Resource optimization (opens in a new tab) through Google Cloud technologies is the practice of using only the compute, storage, and network resources a workload truly needs. When virtual machines (VMs) are right-sized, idle resources are removed, and data lifecycles are managed, data centers consume less electricity and produce fewer emissions. This approach also lowers cost and improves performance, so higher efficiency means lower environmental impact.
An efficient cloud foundation starts with modern Google Cloud data centers that use less energy than typical on-premises rooms. Migrating on-premises workloads to Google Cloud can reduce energy usage and associated emissions by 1.4 to 2 times compared with typical on-premises deployments. Cloud data centers are custom-designed for high power usage effectiveness (PUE), which measures how efficiently a facility uses energy for computing versus cooling and power distribution. Older on-premises data centers often lack the scale to justify investments in advanced cooling and power systems.
Shared responsibility for sustainable use divides work between Google and the customer in a shared-fate partnership to reduce the footprint of the whole ecosystem. Google is responsible for sustainability of Google Cloud, which covers energy efficiency and water stewardship of data centers, infrastructure, and core services through renewable energy, climate-conscious cooling, and hardware optimization. The customer is responsible for sustainability in the cloud, which means optimizing workloads by right-sizing resources, using serverless services that scale to zero, and managing data lifecycles effectively.
Intelligent workload management matches provisioned capacity to actual demand so compute does not run idle and waste energy. Active Assist provides right-sizing recommenders, such as machine type recommendations for Compute Engine VMs, that suggest a smaller or more efficient machine type based on utilization. Aggressive autoscaling expands capacity during peaks and shrinks it afterward, while serverless services scale to zero when no requests arrive, so no energy is spent between events. Strong security and reliability also support efficiency because they help prevent spikes from denial of service (DoS) attacks, breaches, and lengthy cleanup operations.
Targeted actions turn these ideas into verified reductions in carbon and cost:
Data and storage optimization reduces the energy cost of keeping, copying, and moving data across regions. Cloud Storage lifecycle rules transition infrequently accessed data to cold storage classes such as Nearline, Coldline, or Archive, or delete data that is no longer needed. For analytics workloads, compressed columnar formats like Parquet or optimized Avro use far less disk space and read more efficiently than row-based formats like JSON or CSV, which lowers energy for compute and input/output operations. Regular audits of persistent disks attached to Compute Engine instances, with deletion of old and unused snapshots, further prevent over-provisioned storage.
Location choices also control network energy and emissions, so data should be stored in regions with higher carbon-free energy (CFE) or lower grid carbon intensity. Replication across regions should be used only to meet mandatory disaster recovery (DR) or high-availability (HA) requirements because cross-region copying increases energy use. Compute-intensive work such as AI training and BigQuery processing should run in the same region as the data source, and large transfers should use Storage Transfer Service or data-sharing APIs instead of mass data dumps.
The continuous improvement sequence for carbon and cost is an ordered maturity path from baseline measurement to embedded GreenOps governance. BigQuery (opens in a new tab) sits at the center of this flow, because Carbon Footprint data flows from Google Cloud services to BigQuery, where teams correlate carbon data with billing and usage data and build custom reports in tools such as Looker.
A four-phase process flow showing the continuous improvement sequence: establish a baseline, identify carbon hotspots, implement targeted optimization, and institutionalize GreenOps, with BigQuery carbon exports looping back to the baseline.
roles/billing.carbonViewer) to FinOps and platform teams, organize folders, projects, tags, and labels for correct attribution, automate export to BigQuery, and define carbon intensity as kilograms of CO2 equivalent per customer, transaction, or revenue unit.The two accounting methods compare as follows:
| Accounting method | What it measures | When to use it |
|---|---|---|
| Location-based | Emissions factor of the local electrical grid | To understand the physical-grid impact of a workload in a region |
| Market-based | Impact after Google purchases of carbon-free energy (CFE) | To understand the carbon benefit from Google CFE investments |
Exam tip: Cost often trends with carbon but is only an initial indicator, because a high-cost region with a high CFE score can have high cost with low market-based emissions, so always validate with both location-based and market-based metrics.
AI-driven recommendations for sustainability turn vast datasets into actions that reduce waste beyond a single workload. AI enables a shift from reactive compliance to proactive optimization across operational efficiency, transparency and risk, and value and growth. Operational efficiency includes inventory management, supply chain optimization, and intelligent energy management, while transparency and risk includes supply chain visibility, regulatory compliance, and climate risk modeling.
Google Cloud offers products that support these outcomes:
Carbon Footprint (opens in a new tab) is the central Google Cloud tool that turns broad sustainability goals into actions teams can track. Data flows from Google Cloud services to the Carbon Footprint dashboard and to exports, where teams monitor impact, find hotspots, and guide operational decisions. Every resource created in Google Cloud has an associated carbon footprint, so measurement must become part of the normal operational feedback loop.
What Carbon Footprint measures is gross greenhouse gas emissions from electricity associated with the use of covered Google Cloud services. The dashboard aggregates emissions at the project level and lets teams break the data down by project, region, and service to see which workloads drive the most impact. This breakdown makes it possible to prioritize the largest emitters first and to track whether fixes actually reduce emissions over time.
How emissions are calculated is through a transparent, auditable, and globally aligned methodology based primarily on the Greenhouse Gas (GHG) Protocol for carbon reporting and accounting. The methodology covers data sources, Scope-3 inclusions, and a customer allocation model that attributes shared infrastructure emissions to each customer. The report uses two accounting methods together, so teams can see both physical grid impact and the benefit of Google's clean energy purchases.
| Accounting method | What it measures | What it reveals |
|---|---|---|
| Location-based | Emissions factor of the local electrical grid | The physical-grid impact where the workload ran |
| Market-based | Impact after Google's purchases of carbon-free energy (CFE) | The carbon benefit from Google's CFE investments |
Setting up measurement and reporting means establishing a baseline that teams can trust and reuse. A well-defined resource hierarchy of folders and projects plus accurate tags and labels is essential, because Carbon Footprint data is aggregated at the project level and correct attribution depends on that structure. Once the structure is correct, teams configure access, exports, and metrics in order:
roles/billing.carbonViewer) on the billing account.Monthly BigQuery exports can feed directly into existing FinOps dashboards such as Looker, so carbon and cost appear in a single review interface.
Finding carbon hotspots means analyzing granular Carbon Footprint data to identify the largest sources of emissions and verify the results of fixes. Teams slice the dashboard and exported data to prioritize action, then combine carbon data with billing data to make decisions that balance sustainability and cost:
Verified, auditable Carbon Footprint data in BigQuery also supports formal environmental, social, and governance (ESG) disclosures for transparency and compliance. Over time, a GreenOps function uses this combined carbon and cost view to set accountability for reduction targets and track carbon efficiency against spending.
Exam tip: Cost and carbon often trend together, but cost is only an initial indicator — in a high-cost region with a high Carbon-Free Energy score, a workload can have high cost but low market-based emissions.
roles/billing.carbonViewer), with exports to BigQuery for correlating carbon, cost, and usage data.Sustainability of Google Cloud is Google's responsibility for energy efficiency and water stewardship of data centers, infrastructure and core services through renewable energy, climate-conscious cooling and hardware optimization. Sustainability in the cloud is the customer's responsibility for energy efficiency of workloads, including right-sizing resources, using serverless services that scale to zero and managing data lifecycles.
Google Cloud Carbon Footprint displays gross greenhouse gas emissions from electricity associated with covered Google Cloud services, aggregated at the project level and broken down by project, region and service. Location-based accounting reflects the emissions factor of the local electrical grid to show physical-grid impact, while market-based accounting reflects impact after Google's carbon-free energy purchases to show the benefit of those investments.
Migrating on-premises workloads to Google Cloud reduces energy use and associated emissions by 1.4 to 2 times compared with typical on-premises deployments. The savings come from custom-designed data centers for high power usage effectiveness (PUE), which measures how efficiently a facility uses energy for computing versus cooling and power distribution, while older on-premises rooms often lack the scale to justify advanced cooling and power systems.
A retail organization is modernizing its application infrastructure on Google Cloud to reduce its environmental footprint. The engineering team needs to optimize two primary workloads:
Which cloud-native strategy should the organization implement to minimize energy waste and reduce carbon emissions?
Deploy the microservices on Cloud Run to automatically scale to zero when idle, and use carbon-aware scheduling based on BigQuery regional energy datasets for the batch jobs
Migrate compute-intensive batch workloads from specialized TPU or GPU accelerators to standard CPU-only virtual machines to lower the nominal power draw per server
Maintain an active-passive disaster recovery architecture with pre-provisioned Compute Engine instances running continuously in a secondary region
Configure vertical scaling on fixed-size virtual machines to accommodate peak traffic loads instead of implementing horizontal autoscaling
Last updated