179 questions
No questions match those filters.
← All questions
System DesignHard· Multi-tenancy· Capacity Planning· Cost Optimization
You operate two tenants against one index: tenant A run...
This is one of the questions in the full AI/ML interview bank. Pro unlocks all 1789 questions; Premium includes the same bank plus the highest daily Practice limit.
See plansDesign should distinguish between fixed costs (memory/GPU capacity) and variable costs (compute per query). For high-volume tenants like A, dedicated hardware is efficient because fixed costs are amortized over high volume. For low-volume, bursty tenants like B, reserved hardware leads to high idle costs; shared or serverless infrastructure is more appropriate despite the potential for cold-start latency.