After merging your PR, Conbench analyzed the 4 benchmarking runs that have been run so far on merge-commit ed36107.

Benchmarks with performance regressions

There were 0 possible performance regressions, according to the lookback z-score method.

All benchmark runs analyzed:

• Commit Run on test-mac-arm at 2025-05-21 09:04:41Z
• Commit Run on arm64-t4g-2xlarge-linux at 2025-05-21 06:17:24Z
• Commit Run on amd64-c6a-4xlarge-linux at 2025-05-21 02:54:08Z
• Commit Run on amd64-m5-4xlarge-linux at 2025-05-21 04:13:40Z

Unstable benchmarks with performance regressions

The following benchmark results indicate a possible performance regression, but are known to sometimes produce false positives when applying the lookback z-score analysis.

• Commit Run on arm64-t4g-2xlarge-linux at 2025-05-21 06:17:24Z

  • ResolveManyUInt16Sorted (C++) with params=chunked_array_length:65535/num_chunks:100/num_indices:32764, source=cpp-micro, suite=arrow-chunk-resolver-benchmark
  • ResolveManyUInt32Random (C++) with params=chunked_array_length:65535/num_chunks:10000/num_indices:32764, source=cpp-micro, suite=arrow-chunk-resolver-benchmark

• Commit Run on amd64-c6a-4xlarge-linux at 2025-05-21 02:54:08Z

  • MemoryBandwidth (C++) with params=/33554432/real_time/threads:1, source=cpp-micro, suite=arrow-io-memory-benchmark
  • MemoryBandwidth (C++) with params=/33554432/real_time/threads:2, source=cpp-micro, suite=arrow-io-memory-benchmark