Cloud Digital Leader Fundamental Cloud Concepts Practice Question

Answer analysis

Option-by-option breakdown

For each option: why learners choose it and why it is or isn't the right answer here.

• ✗

  Use committed use discounts to reserve resources

  Why it's wrong here

  Committed use discounts are contractual pricing agreements that lower compute costs in exchange for a 1- or 3-year resource commitment, but they do not change the energy efficiency of the underlying hardware, the grid carbon intensity, or how often the workload actually runs. In fact, a CUD may encourage you to keep idle resources running to realize the discount, potentially increasing wasted energy and emissions. Therefore, while attractive for cost control, they are not a carbon-reduction lever.

• ✓

  Shift non-urgent compute loads to times when low-carbon energy is available

  Why this is correct

  Carbon-intelligent load shifting exploits the fact that the carbon intensity of grid electricity fluctuates throughout the day as renewable sources like wind and solar come online and offline. By scheduling batch jobs, data pipelines, or other latency-tolerant workloads during hours when the regional energy mix is cleanest, the same compute tasks produce significantly less CO2 without any efficiency changes. This is an operational, time-based strategy that directly targets the emissions-per-kilowatt-hour of the power consumed.

• ✓

  Choose regions with lower carbon intensity

  Why this is correct

  Every cloud region draws its electricity from a different energy portfolio — some are dominated by hydro or wind, while others still rely on coal or heavy natural gas. Selecting a region with a lower published carbon intensity (for example, Google Cloud's region carbon data) means the identical workload will emit fewer metric tons of carbon because the underlying grid is cleaner. This is a geo-location decision, and it must be balanced with latency, data residency, and cost, but it is a direct way to shrink your cloud carbon footprint.

• ✗

  Provision VMs with GPUs for general workloads

  Why it's wrong here

  GPUs are purpose-built for highly parallel workloads such as deep learning, scientific simulation, and 3D rendering, and they draw significant power, often consuming hundreds of watts per board. Using them for general-purpose tasks like simple web serving or lightweight application backends wastes that energy because the parallelism is unused and the hardware sits mostly idle, driving up both energy consumption and embodied manufacturing impact. As a result, over-provisioning with GPUs for general workloads increases carbon emissions rather than mitigating them.

• ✓

  Use the Cloud Carbon Footprint tool to track emissions

  Why this is correct

  The Cloud Carbon Footprint tool in Google Cloud generates emissions estimates based on your actual resource usage, disaggregated by project, service, and region, giving you the data needed to identify the highest-emission workloads. It is the 'measure-first' component of a cloud sustainability strategy, because without accurate baselines you cannot evaluate the effectiveness of load shifting, region moves, or hardware changes. While measurement alone does not reduce emissions, it is a required prerequisite for purpose-built optimization.

Go deeper

Related to this question