CPU Design · All levels

FPU and Vector Units: Interview Drills

Interview Drills for FPU and Vector Units.

Interview drills

Interview Drills for FPU and Vector Units centers on FP/vector utilization, latency overlap efficiency, and denormal handling penalties. Tie every claim to a measurable artifact and an owner-controlled action.

diagram
PROMPT
You observe FP/vector utilization, latency overlap efficiency, and denormal handling penalties on FPU and Vector Units. Walk through root cause and release decision.

STRONG ANSWER
1. Names failing workload and first stage loss.
2. Explains mechanism: Floating-point and vector pipelines have distinct latencies and lane widths; scheduler and compiler coordination is required to hide long operations while avoiding port oversubscription.
3. Requests proving artifact: vector lane utilization map, FP latency histogram, and exception handling trace
4. Proposes bounded fix + owner + rollback-safe validation.

WEAK ANSWER
Suggests generic optimization ideas without mechanism proof or owner accountability.

Decision matrix

diagram
CPU EVIDENCE MATRIX - FPU and Vector Units

+---------------------------+--------------------------------+--------------------------------+---------------------------+
| Evidence                  | Tells you                      | Does not prove                 | Next action               |
+---------------------------+--------------------------------+--------------------------------+---------------------------+
| CPI + top-down stack      | broad pressure domain          | exact root mechanism           | inspect first failing stage |
| PMU event timeline        | temporal onset and persistence | causality by itself            | pair with trace and config lock |
| pipeline occupancy trace  | bubble origin and spread       | multicore/system interactions  | correlate with LLC/NoC data |
| cache/TLB/coherence logs  | memory and translation health  | scheduler fairness             | inspect issue/port behavior |
| thermal + power telemetry | silicon operating envelope     | architectural correctness      | validate bounded fixes at same corners |
+---------------------------+--------------------------------+--------------------------------+---------------------------+

CPU deep dive

Execution throughput depends on port balance, bypass quality, and realistic instruction mix assumptions.

Concept diagram

diagram
EXECUTION DATAPATH

issue -> ALU/FPU/vector/LSQ ports -> writeback -> retire

Metric graph

diagram
EXECUTION LOSS DRIVERS

port conflicts      █████
bypass hazards      ████
LSQ ordering stalls ███

Reports and artifacts

  • port pressure heatmap

  • pipeline hazard report

  • ALU/FPU/vector utilization split

  • LSQ ordering diagnostics

Mini case study

A compiler scheduling update over-concentrated uops on one port class, reducing effective multi-issue throughput.

Debug branches

  • Map instruction classes to port availability

  • Validate forwarding depth against dependency chains

  • Inspect LSQ ordering events before widening pipes

Senior review question

Ask: which CPI/latency evidence proves this topic is truly closed beyond synthetic benchmarks?

Key takeaways

  • Always connect microarchitectural counter changes to product workload outcomes.

  • Lock binary, compiler, firmware, and thermal metadata before comparing CPU traces.

Common pitfalls

  • Treating average IPC as sufficient proof while ignoring latency tails and outliers.

  • Applying predictor or prefetch tweaks without first-failing-stage attribution.

  • Declaring closure without reproducible perf, correctness, and power gates.

Interview answer expansion

Strong interview answers for FPU and Vector Units begin with workload and metric framing, then explain mechanism in plain terms: Floating-point and vector pipelines have distinct latencies and lane widths; scheduler and compiler coordination is required to hide long operations while avoiding port oversubscription.

Then propose a measurement plan: CPI stack, branch quality, queue occupancy, cache/TLB behavior, and latency distribution as applicable.

Finally, present one bounded fix plus regression risk. CPU interviews reward explicit tradeoff ownership, not generic tuning advice.