NUMA and Memory Affinity: Debug Playbook

Debug playbook

Debug Playbook for NUMA and Memory Affinity centers on remote memory access ratio, NUMA latency delta, and workload affinity score. 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 - NUMA and Memory Affinity

remote memory access ratio, NUMA latency delta, and workload affinity score 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 - Multicore & System Integration / NUMA and Memory Affinity

1. Symptom
   - Watched metric: remote memory access ratio, NUMA latency delta, and workload affinity score
   - 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: Thread and page placement policy controls whether cores access local or remote memory; poor affinity silently adds latency and coherence overhead to otherwise efficient software.
   - 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: NUMA locality profile, page-placement log, and latency percentile report
   - Required owners: system software owner, platform architect, performance engineer
   - Final decision: ship, bounded rollout, rollback, or escalate

CPU deep dive

Scaling across cores is limited by coherence and interconnect behavior before compute saturation on many workloads.

Concept diagram

MULTICORE SYSTEM VIEW

cores + private caches <-> LLC slices <-> interconnect <-> memory controllers

Metric graph

SCALING EFFICIENCY

ideal scaling        ███████████
observed under load  ███████
after policy tuning  █████████

Reports and artifacts

• coherence traffic matrix

• NoC congestion map

• NUMA locality profile

• synchronization contention report

Mini case study

A lock-heavy service regressed at higher core counts because coherence invalidations and NoC hotspots dominated.

Debug branches

• Classify traffic as coherence, demand miss, or synchronization

• Measure hotspot links instead of aggregate NoC throughput

• Validate thread and page affinity before hardware changes

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

NUMA and Memory Affinity 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.

Thread and page placement policy controls whether cores access local or remote memory; poor affinity silently adds latency and coherence overhead to otherwise efficient software. 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 remote memory access ratio, NUMA latency delta, and workload affinity score 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 NUMA locality profile, page-placement log, and latency percentile report.

Multicore scaling is governed by coherence traffic, interconnect fairness, and memory placement discipline. 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.