[Perf] Linux/x64: 1 Improvement on 6/2/2026 10:24:00 PM +00:00 #74255

Activity

1. performanceautofiler

   added

   arch-x64

   branch-refs/heads/main

   compilationmode-tiered

   kind-micro

   os-linux

   perf-improvement

   runkind-micro

   runtime-coreclr

   untriaged

   on Jun 9, 2026
2. LoopedBard3 commented on Jun 9, 2026

   

   LoopedBard3

   on Jun 9, 2026

   Member

   More actions

   🔍 Automated Triage Analysis

   Summary: All reported changes appear to be measurement noise. The single test (System.Linq.Tests.Perf_Enumerable.SingleWithPredicate_FirstElementMatches) shows a -65% raw delta but historical CV is 76.6% with extreme variance from spike clusters, placing the change well within normal fluctuation (z=-0.94, adaptive threshold 153%). Finding confidence 4/5 (strong historical data, 291 runs), improvement confidence 1/5 (classified as noise).

   Finding Confidence: 4/5 (analysis accuracy)
   Improvement Confidence: 1/5 (likelihood of true improvement)
   ⚠️ Assessment: Likely noise - recommend closing or monitoring for recurrence.

   🤖 Proposed Next Actions: Likely noise — close & label — 3 proposed (rule R1_all_noise — see Full Analysis for the table).

   📊 Full Analysis (click to expand)

   Summary: All reported changes appear to be measurement noise. The single test (System.Linq.Tests.Perf_Enumerable.SingleWithPredicate_FirstElementMatches) shows a -65% raw delta but historical CV is 76.6% with extreme variance from spike clusters, placing the change well within normal fluctuation (z=-0.94, adaptive threshold 153%). Finding confidence 4/5 (strong historical data, 291 runs), improvement confidence 1/5 (classified as noise).

   Issue Overview

   • Issue: [Perf] Linux/x64: 1 Improvement on 6/2/2026 10:24:00 PM +00:00 #74255 - [Perf] Linux/x64: 1 Improvement on 6/2/2026 10:24:00 PM +00:00
   • Type: Performance Improvement
   • Date: 2026-06-02 | Queue: Ubuntu.2204.Amd64.Viper.Perf | OS/Arch: linux/x64
   • Commits: c7d89359fb2a → ac695692911 (28 commits)
   • Labels: untriaged, perf-improvement, arch-x64, os-linux, branch-refs/heads/main, runtime-coreclr, kind-micro, compilationmode-tiered, runkind-micro
   • Runtime Stack: Desktop + CoreCLR (implicit)
   • Historical Data: Available (291 prior runs, 322 total examined)

   Test Analysis

   Test Name	Baseline	Compare	Δ%	μ	σ	n	z	Thresh%	Noise?	Assessment
   SingleWithPredicate_FirstElementMatches(input: IEnumerable)	96.44 ns	33.36 ns	-65.4%	117.80	90.18	291	-0.94	153.1%	Y	Within extreme historical variance; bimodal clusters

   Legend: μ=mean, σ=stddev, n=runs, z=z-score, Thresh%=adaptive threshold

   Finding Confidence: 4/5 - Strong historical data with 291 prior runs providing excellent coverage. Data quality is high despite the extreme variance.

   Candidate Stack Relevance

   Not applicable — all tests classified as noise; no likely cause attribution needed.

   Noise Classification Details

   System.Linq.Tests.Perf_Enumerable.SingleWithPredicate_FirstElementMatches(input: IEnumerable)

   • Delta: 96.44 ns → 33.36 ns (-65.4%)
   • Historical Context: μ=117.80 ns (σ=90.18, n=291); z-score=-0.94; adaptive threshold=153.1%; CV=76.6%
   • Classification: Noise
   • Rationale: Despite a large raw delta of -65.4%, the absolute z-score (0.94) is well below the significance threshold of 2.0, and the percent change is far below the adaptive threshold of 153.1%. The extreme historical CV (76.6%) is driven by a transient spike cluster (values 499–581 ns around indices 52–64) that inflates stddev dramatically. The compare values (~32–36 ns and ~56 ns) are bimodal but remain within 1 SD of the historical mean. This test has inherent high variance that makes the current delta indistinguishable from noise.
   • Change Pattern: Unknown (no clear step change or drift detected)

   Open Questions

   • Historical Data: Extremely high variance (CV=76.6%) makes it difficult to detect genuine improvements with this test
   • Spike Cluster: Historical series contains a transient spike cluster (values 499–581 ns) around indices 52–64 that dramatically inflates stddev; if those outliers have been resolved, re-baselining would tighten the threshold
   • Bimodal Compare: After-values show two clusters (~32–36 ns and ~56 ns), suggesting possible environment switching or flaky benchmark behavior
   • Verification Needed: Not required — classified as noise with high confidence

   Recommended Actions

   1. Re-baseline consideration - If the historical spike cluster (499–581 ns) represents a resolved transient issue, re-baselining the test history would reduce σ and make future genuine improvements detectable
   2. Investigate bimodality - The bimodal compare values (~32–36 ns vs ~56 ns) may indicate environment-dependent behavior worth investigating for benchmark stability
   3. No further triage needed - This change is noise under current statistical thresholds; no attribution or bisection warranted

   🤖 Proposed Next Actions

   Rule: R1_all_noise — Likely noise — close & label.

   All reported improvements were classified as noise.

   #	Action	Status	Details
   1	Add label	🕒 Planned	noise
   2	Remove label	🕒 Planned	untriaged
   3	Close issue	🕒 Planned	reason not_planned

   ---

   Generated by PerfTriageAgent on 2026-06-09 11:13 UTC
3. LoopedBard3

   added

   Noise

   on Jun 9, 2026
4. AndyAyersMS commented on Jun 9, 2026

   

   AndyAyersMS

   on Jun 9, 2026

   Member

   More actions

   dotnet/runtime#128513
5. AndyAyersMS

   closed this as completedon Jun 9, 2026
6. LoopedBard3

   added

   Gold Star (Bot)When the triage bot does good.When the triage bot does good.

   and removed

   Gold Star (Bot)When the triage bot does good.When the triage bot does good.

   on Jun 11, 2026