M4.8 module quiz
Ten questions. 80% to pass (8 of 10). Open book, unlimited retakes.
Answers are collapsed under each question. Answer first, then check.
Q1
Two carbon estimates in one report use 100 W and 50 W for the same instance type. The right fix is:
A. Average the two published vendor figures
B. Use the higher of the two figures, to be conservative
C. Report a range and no point estimate at all
D. Pick one, state it explicitly, and apply it consistently
Show answer
Correct: D. All downstream figures differ by exactly the ratio, and a report whose internal assumptions disagree is worse than no report because the inconsistency is the first thing a reviewer finds.
Q2
AWS publishes per-instance wattage:
A. In the pricing API, per instance type
B. In the provider’s sustainability dashboard
C. It does not; any figure is an estimate
D. Per region, in the carbon intensity feed
Show answer
Correct: C. Which is why the assumption has to be stated rather than presented as a fact, and why two teams using different sources produce different numbers legitimately.
Q3
Region carbon intensity varies primarily with:
A. The generation mix of the local grid
B. Distance of the region from the equator
C. The age of the data centre buildings
D. Which cloud provider operates it
Show answer
Correct: A. Which is why the same workload in two regions of the same provider can differ several-fold in emissions with identical cost.
Q4
Carbon-aware computing shifts workloads by:
A. The cloud provider that you run it on
B. Time, geography, or throttling to batch
C. The instance family it runs on
D. The storage class it writes to
Show answer
Correct: B. All three require the workload to be deferrable or relocatable, which is why the candidate set is much narrower than the total estate.
Q5
Cost and carbon align when:
A. Always, on absolutely every single optimization action
B. Never; they are separate concerns
C. Eliminating waste, since unused capacity emits and costs
D. Only in renewable-powered regions
Show answer
Correct: C. Waste elimination is the case where the two objectives point the same way. They diverge when the cleanest region is not the cheapest, which is a real trade rather than a free win.
Q6
Scheduling reduces carbon:
A. Not at all, since the underlying capacity still exists regardless of it
B. Only for GPU workloads, which dominate draw
C. Only in regions with a renewable grid mix
D. Proportionally to the hours removed, subject to the grid mix in those hours
Show answer
Correct: D. Which is why a schedule can produce a larger carbon reduction than cost reduction, or vice versa, depending on when the removed hours fall on the grid’s daily curve.
Q7
The double-lever framing means:
A. Cost and carbon respond identically to exactly the same action taken
B. One action can be justified on two grounds, though the magnitudes differ
C. Two separate actions are always required for each
D. Carbon has to be measured twice to be trusted
Show answer
Correct: B. Assuming the same percentage applies to both ignores grid-intensity variation, which is the most common error in a combined business case.
Q8
A carbon report to leadership should lead with:
A. The full methodology behind the figures
B. The full set of per-instance wattage assumptions used
C. The trend and the actions available, with assumptions stated
D. Regional comparisons across the estate
Show answer
Correct: C. Methodology belongs in an appendix. Leading with it invites a debate about the estimate rather than a decision about the action.
Q9
Reporting carbon alongside cost is most credible when:
A. The two are combined into one blended score
B. Only carbon is reported, with cost implied
C. Carbon is expressed in dollars for a much easier comparison
D. They are reported separately with their own trends and assumptions
Show answer
Correct: D. A combined score hides which lever moved, and a carbon figure expressed in dollars invites an argument about the conversion rate rather than the emissions.
Q10
The most common carbon reporting pitfall is:
A. Assumption drift between sections of the same report
B. Over-reporting the total headline figure upward
C. Reporting it far too frequently
D. Using the wrong unit throughout
Show answer
Correct: A. Year figures, driver numbers and wattage assumptions that differ across three sections of one document are the failure that costs the report its credibility.
What’s next
Back to Sustainability + carbon.