Performance Counters (PMC): Debug Playbook

Debug playbook

Debug Playbook for Performance Counters (PMC) centers on counter fidelity, sampling overhead, and triage turnaround time. Tie every claim to a measurable artifact and an owner-controlled action.

1. Freeze workload seed, binary, compiler, firmware, and thermal setup.

2. Find first persistent stage loss in timeline.

3. Build one reduced reproducer for dominant hypothesis.

4. Patch minimal fix with explicit rollback gate.

5. Re-run full correctness + performance + power matrix.

Debug decision tree

ROOT-CAUSE TREE - Performance Counters (PMC)

counter fidelity, sampling overhead, and triage turnaround time regressed
        |
  reproducible on fixed seed?
      /               \
    no                 yes
    |                   |
env/tool drift      first failing stage?
                    /        |        \
                front-end   execute   memory/system
                   |          |            |
              fetch/decode   port/ROB   cache/TLB/NoC

Stop at first confirmed mechanism, then patch with owner accountability.

Review memo template

CPU DESIGN REVIEW MEMO - Physical Design, Perf & Bring-up / Performance Counters (PMC)

1. Symptom
   - Watched metric: counter fidelity, sampling overhead, and triage turnaround time
   - Failing workload slice: <name>
   - First failing stage: <fetch/decode/rename/execute/memory/system>
   - Revision tags: <binary/compiler/firmware/uarch stepping>

2. Mechanism hypothesis
   - Primary mechanism: PMCs expose stall classes and throughput limits; the value comes from event taxonomy quality, synchronized sampling, and disciplined correlation against workload traces.
   - Competing hypotheses: <front-end, scheduler, memory, coherence, physical limits>
   - Missing evidence: <counter snapshot, trace, topology/thermal map>

3. Proposed action
   - Minimal reversible fix: <uarch policy/compiler/runtime/config>
   - Expected movement: <IPC/CPI/latency tail/perf-per-watt>
   - Regression risk: correctness, power, thermal, software compatibility

4. Signoff
   - Re-run artifact: PMU event map, counter correlation notebook, and anomaly triage report
   - Required owners: silicon performance lead, firmware owner, tooling owner
   - Final decision: ship, bounded rollout, rollback, or escalate

CPU deep dive

Physical closure and observability planning determine whether CPU architecture wins survive first silicon.

Concept diagram

CPU SILICON CLOSURE

core/LLC floorplan -> clock/power domains -> PMCs/observability -> bring-up

Metric graph

CLOSURE RISK MIX

timing margin risk   █████
thermal hotspots     ████
bring-up blockers    ███

Reports and artifacts

• floorplan congestion map

• timing closure summary

• IR/thermal transient report

• bring-up milestone tracker

Mini case study

A floorplan change improved routing congestion but created thermal clustering that forced frequency throttling in sustained tests.

Debug branches

• Trace critical paths to physical regions and domain crossings

• Run dynamic IR and thermal checks on burst workloads

• Use PMCs and bring-up logs to correlate silicon symptoms to design intent

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.

Principal CPU review addendum

Performance Counters (PMC) should be treated as a system behavior, not an isolated block definition. In a shipping CPU core, ISA intent, front-end delivery, speculation depth, scheduler behavior, memory translation, coherence traffic, and physical limits all interact before software observes final IPC or CPI.

PMCs expose stall classes and throughput limits; the value comes from event taxonomy quality, synchronized sampling, and disciplined correlation against workload traces. CPU teams pay for repeated inefficiency: one extra bubble, one wrong target, one port conflict, or one translation miss pattern can replicate across billions of instructions and dominate product-level latency and energy.

Use counter fidelity, sampling overhead, and triage turnaround time as an investigation start point, not as the conclusion. A counter movement only becomes actionable when paired with workload phase tags, PMU event context, a controlled repro, and artifact evidence such as PMU event map, counter correlation notebook, and anomaly triage report.

CPU product success depends on physical closure and observability being designed into architecture choices early. Senior review quality comes from proving the full chain: workload request -> microarchitectural response -> measured bottleneck -> smallest owner fix -> regression-safe validation.

Review discipline should force a causal chain: workload shape -> front-end/speculation behavior -> execution/memory pressure -> retire efficiency -> product impact. That chain keeps CPU decisions evidence-driven and owner-accountable.