Practice cost vs. performance vocabulary: latency-cost trade-offs, spot instances, performance SLOs vs. cost, per-API-call budgets, and optimization language.
0 / 5 completed
1 / 5
'We can reduce cost by 30% by accepting 50ms more ___.' Which noun is the performance variable?
'Latency' is the performance variable being traded off against cost. Accepting slightly higher latency (e.g., by using a cheaper instance type) can yield significant cost savings.
2 / 5
'The ___ instance saves 70% but has interruption risk.' Which adjective describes discounted preemptible compute?
'Spot instances' (AWS) or 'preemptible VMs' (GCP) offer up to 70–90% discounts but can be terminated with 2 minutes notice. They're cost-effective for fault-tolerant or batch workloads.
3 / 5
'The ___ between performance SLO and cost' — which noun describes a decision with competing objectives?
'Trade-off' is the standard business and engineering term for decisions where improving one dimension (performance) increases another (cost). Managing trade-offs is central to platform economics.
4 / 5
'We set a ___ budget per API call.' Which adjective describes money allocated per unit of computation?
'Cost budget per API call' defines the maximum acceptable cost to serve each API request. This drives architecture decisions: caching, efficient queries, right-sizing — to stay within the budget.
5 / 5
What does it mean to 'right-size' a cloud resource?
'Right-sizing' means matching the compute, memory, and storage allocation to actual workload needs. Over-provisioned resources waste money; under-provisioned resources hurt performance. Right-sizing optimizes both.