After merging your PR, Conbench analyzed the 4 benchmarking runs that have been run so far on merge-commit f10c93c.

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 2026-07-02 06:06:37Z
• Commit Run on arm64-t4g-2xlarge-linux at 2026-07-01 22:42:24Z
• Commit Run on amd64-m5-4xlarge-linux at 2026-07-01 23:28:06Z
• Commit Run on amd64-c6a-4xlarge-linux at 2026-07-01 22:16:44Z

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 test-mac-arm at 2026-07-02 06:06:37Z

  • BM_UnpackUint8 (C++) with params=DynamicUnaligned/2/128, source=cpp-micro, suite=arrow-bpacking-benchmark
  • BM_UnpackUint8 (C++) with params=NeonUnaligned/2/128, source=cpp-micro, suite=arrow-bpacking-benchmark

• Commit Run on amd64-m5-4xlarge-linux at 2026-07-01 23:28:06Z

  • TableToTensorSimple (C++) with params=<Int64Type, true>/size:32768/num_columns:3, source=cpp-micro, suite=arrow-tensor-benchmark
  • TableToTensorSimple (C++) with params=<Int16Type, true>/size:32768/num_columns:3, source=cpp-micro, suite=arrow-tensor-benchmark
  • TableToTensorSimple (C++) with params=<Int8Type, true>/size:32768/num_columns:30, source=cpp-micro, suite=arrow-tensor-benchmark

• Commit Run on amd64-c6a-4xlarge-linux at 2026-07-01 22:16:44Z

  • BatchToTensorSimple (C++) with params=/size:524288/num_columns:3, source=cpp-micro, suite=arrow-tensor-benchmark
  • BatchToTensorSimple (C++) with params=/size:32768/num_columns:30, source=cpp-micro, suite=arrow-tensor-benchmark
  • BatchToTensorSimple (C++) with params=/size:524288/num_columns:300, source=cpp-micro, suite=arrow-tensor-benchmark
  • BatchToTensorSimple (C++) with params=/size:524288/num_columns:30, source=cpp-micro, suite=arrow-tensor-benchmark
  • BatchToTensorSimple (C++) with params=/size:524288/num_columns:30, source=cpp-micro, suite=arrow-tensor-benchmark
  • BM_ByteStreamSplitDecode_FLBA_Generic (C++) with params=<2>/1024, source=cpp-micro, suite=parquet-encoding-benchmark
  • BatchToTensorSimple (C++) with params=/size:32768/num_columns:3, source=cpp-micro, suite=arrow-tensor-benchmark