The Impact of Using GenAI on Energy Consumption and CO2 Emissions CT-GENAI Practice Questions (Page 3)

Questions 11–15

1. 11expert · hard

   A QA team uses GenAI to generate test cases for a legacy system. The GenAI model produces test cases that find 10% more defects than manual testing, but the GenAI generation consumes 200 kWh per release. Manual testing consumes 5 kWh per release (for test execution) but requires 50 person-hours. The company's sustainability policy prioritizes reducing CO2 emissions. What is the most defensible conclusion?

   • AGenAI is never justified because 200 kWh is always more than 5 kWh
   • BManual testing is always more sustainable because it uses less compute energy
   • CGenAI is justified if the 10% additional defect detection prevents production failures whose remediation energy exceeds the 200 kWh
   • DThe comparison is impossible because energy and person-hours cannot be compared
   Reveal Solution Discussion Select an answer first
2. 12application · medium · select all that apply

   A test team wants to reduce the CO2 emissions of their GenAI-based test generation pipeline. The pipeline runs on a cloud provider. Select all strategies that can reduce emissions without reducing test coverage.

   • AUse a smaller, fine-tuned model that produces equivalent test outputs
   • BMove the workload to a region with lower carbon intensity
   • CCache and reuse generated tests when the application under test has not changed
   • DIncrease the number of generated test cases to improve coverage
   • ESchedule inference jobs during periods when the grid's renewable energy share is highest
   Check answers Discussion Select an answer first
4. 13application · hard

   A QA team is evaluating whether to use GenAI to generate test data or use a traditional scripted data generator. The GenAI approach consumes 5 kWh per test cycle. The scripted approach consumes 0.5 kWh per cycle but requires 10 hours of developer time to maintain per release. The team runs 100 test cycles per year. What is the most important factor in deciding which approach is more environmentally sustainable?

   • AThe programming language used for the scripted generator
   • BThe number of test cases generated per cycle
   • CThe age of the GenAI model
   • DThe energy cost of the developer's workstation and commuting during the 10 hours of maintenance
   Reveal Solution Discussion Select an answer first
5. 14application · easy

   A test team uses a GenAI model to generate test data for a mobile app. The team notices that the model generates 30% more data than needed, and the excess is discarded. The team wants to reduce energy consumption. What is the most effective approach?

   • AIncrease the model's max tokens setting to allow more complete responses
   • BRun the generation twice and keep the better result
   • CAdjust the prompt to request exactly the required number of data records
   • DUse a larger model to improve the quality of the generated data
   Reveal Solution Discussion Select an answer first
6. 15application · medium

   A company's test automation pipeline uses GenAI to generate test scripts. The pipeline runs on a cloud provider that offers a 'carbon-aware' scheduling feature. This feature delays jobs until the grid's carbon intensity drops below a threshold. The team is concerned that delaying jobs may cause test results to be delivered too late for the release deadline. What is the best way to balance emissions reduction and delivery deadlines?

   • ADisable carbon-aware scheduling to guarantee on-time delivery
   • BRun the job only when carbon intensity is below the threshold, regardless of deadline
   • CAlways run the job immediately and purchase carbon offsets
   • DUse carbon-aware scheduling with a maximum delay threshold that still meets the release deadline
   Reveal Solution Discussion Select an answer first