🤖 Update Benchmark Results

[thomhurst]

Automated Benchmark Update

This PR updates the benchmark documentation with the latest results from the Speed Comparison workflow.

Benchmarks Produced

Individual benchmark artifacts are available for download:

• benchmark-DataDrivenTests
• benchmark-AsyncTests
• benchmark-ScaleTests
• benchmark-MatrixTests
• benchmark-MassiveParallelTests
• benchmark-SetupTeardownTests
• benchmark-BuildTime
• benchmark-summary (aggregated overview)

Runtime Benchmarks

• AsyncTests
• DataDrivenTests
• MassiveParallelTests
• MatrixTests
• ScaleTests
• SetupTeardownTests

Build Benchmarks

• BuildTime
• 2026-05-27
• Ubuntu Latest • .NET SDK 10.0.300

Changes

• Updated benchmark data in docs/static/benchmarks/latest.json
• Updated historical trends in docs/static/benchmarks/historical.json
• Regenerated benchmark documentation in docs/docs/benchmarks/index.md
• Updated benchmark summary in docs/static/benchmarks/summary.json

Workflow Run

• Run ID: 26483982160
• Triggered: schedule
• Date:

---

🤖 This PR was automatically created and will be merged automatically once CI checks pass.

[claude]

Benchmark Update Review (2026-05-27)

This is an automated daily benchmark refresh — no logic changes, only updated data and documentation. The automation pipeline looks correct and consistent across all files.

Notable Benchmark Changes

Improvements:

• MassiveParallelTests (TUnit): 653.5ms → 554.3ms (~15% faster) — significant improvement
• AsyncTests (TUnit): 516.1ms → 513.2ms (slight improvement)
• BuildTime (TUnit): 1.697s → 1.586s (~7% faster build)

Regressions to watch:

• DataDrivenTests (TUnit AOT): 24.99ms → 30.90ms (~24% slower) — a notable AOT regression; could be run-to-run noise, but worth monitoring over the next few days
• SetupTeardownTests (TUnit): 521.3ms → 554.1ms (~6% slower) — minor, but worth keeping an eye on
• DataDrivenTests: MSTest (465.59ms) is now faster than TUnit (472.69ms) in this category — TUnit has lost the lead here; this may warrant investigation into what's changed

Structural Observations

The automated update pipeline correctly keeps all these files in sync:

• Per-benchmark docs
• Per-benchmark data
• , ,

No issues with the automation itself. The data format is consistent and timestamps are properly updated.

Recommendation: The AOT regression in DataDrivenTests (+24%) and MSTest overtaking TUnit in DataDrivenTests are worth investigating independently — they may reflect CI infrastructure noise, but they cross meaningful thresholds. Consider adding automated regression alerting to the benchmark workflow to flag these automatically rather than requiring manual review of benchmark PRs.

[claude]

Benchmark Update Review (2026-05-27)

This is an automated daily benchmark refresh — no logic changes, only updated data and documentation. The automation pipeline looks correct and consistent across all files.

Notable Benchmark Changes

Improvements:

• MassiveParallelTests (TUnit): 653.5ms to 554.3ms (~15% faster) — significant improvement
• AsyncTests (TUnit): 516.1ms to 513.2ms (slight improvement)
• BuildTime (TUnit): 1.697s to 1.586s (~7% faster build)

Regressions to watch:

• DataDrivenTests (TUnit AOT): 24.99ms to 30.90ms (~24% slower) — notable AOT regression; could be run-to-run noise, but worth monitoring over the next few days
• SetupTeardownTests (TUnit): 521.3ms to 554.1ms (~6% slower) — minor, but worth keeping an eye on
• DataDrivenTests: MSTest (465.59ms) is now faster than TUnit (472.69ms) in this category — TUnit has lost the lead here; this may warrant investigation

Structural Observations

The automated update pipeline correctly keeps all these files in sync: per-benchmark markdown docs, per-benchmark JSON data, and the aggregated latest/historical/summary JSON files. No issues with the automation itself. Data format is consistent and timestamps are properly updated.

Recommendation: The AOT regression in DataDrivenTests (+24%) and MSTest overtaking TUnit in DataDrivenTests are worth investigating independently. Consider adding automated regression alerting to the benchmark workflow to flag these automatically rather than relying on manual review of benchmark PRs.