After merging your PR, Conbench analyzed the 4 benchmarking runs that have been run so far on merge-commit 22a4496.

Benchmarks with performance regressions

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

All benchmark runs analyzed:

• Commit Run on arm64-t4g-2xlarge-linux at 2024-09-10 17:06:59Z
• Commit Run on test-mac-arm at 2024-09-10 18:37:07Z
• Commit Run on amd64-c6a-4xlarge-linux at 2024-09-10 17:42:12Z
• Commit Run on amd64-m5-4xlarge-linux at 2024-09-10 17:57:35Z

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 2024-09-10 17:06:59Z

  • CheckIndexBoundsUInt32 (C++) with params=65536/1000, source=cpp-micro, suite=arrow-int-util-benchmark
  • TakeFixedSizeBinaryRandomIndicesWithNulls (C++) with params=1048576/1/9, source=cpp-micro, suite=arrow-compute-vector-selection-benchmark

• Commit Run on amd64-c6a-4xlarge-linux at 2024-09-10 17:42:12Z

  • RoundBinaryArrayBenchmark (C++) with params=<RoundBinary, FloatType, RoundMode::DOWN>/size:524288/inverse_null_proportion:100, source=cpp-micro, suite=arrow-compute-scalar-round-benchmark
  • BM_RleEncoding (C++) with params=1024/8, source=cpp-micro, suite=parquet-column-io-benchmark
  • RoundBinaryArrayBenchmark (C++) with params=<RoundBinary, DoubleType, RoundMode::HALF_TOWARDS_ZERO>/size:524288/inverse_null_proportion:0, source=cpp-micro, suite=arrow-compute-scalar-round-benchmark

• Commit Run on amd64-m5-4xlarge-linux at 2024-09-10 17:57:35Z

  • MemoryBandwidth (C++) with params=/33554432/real_time/threads:4, source=cpp-micro, suite=arrow-io-memory-benchmark