perf(parquet): Improve cache locality in BYTE_STREAM_SPLIT decoding

[pchintar]

Which issue does this PR close?

• Closes Improve cache locality in BYTE_STREAM_SPLIT decoding reconstruction #10006 .

Rationale for this change

BYTE_STREAM_SPLIT decoding currently reconstructs values using a nested scalar loop with strided reads across byte streams in:

parquet/src/encodings/decoding/byte_stream_split_decoder.rs

Current logic:

for i in 0..dst.len() / TYPE_SIZE {
    for j in 0..TYPE_SIZE {
        dst[i * TYPE_SIZE + j] = sub_src[i + j * stride];
    }
}

This results in poor cache locality and inefficient memory access patterns during reconstruction of the original value layout.

What changes are included in this PR?

This PR changes the reconstruction logic in join_streams_const and join_streams_variable to process values in contiguous blocks instead of value-by-value scalar iteration.

The updated implementation reads contiguous regions from each byte stream before writing reconstructed values back into the destination buffer.

Example:

for base in (0..values).step_by(BLOCK) {
    let len = (values - base).min(BLOCK);

    for byte_idx in 0..TYPE_SIZE {
        let src_start = byte_idx * stride + base;
        let src_block = &src[src_start..src_start + len];

        for (idx, value) in src_block.iter().copied().enumerate() {
            dst[(base + idx) * TYPE_SIZE + byte_idx] = value;
        }
    }
}

Are these changes tested?

Existing parquet tests pass:

cargo test -p parquet byte_stream_split -- --nocapture
cargo test -p parquet encoding -- --nocapture

Benchmarks from parquet/benches/encoding.rs show considerable improvements for BYTE_STREAM_SPLIT decoding:

cargo bench -p parquet --bench encoding --all-features -- "decoding: dtype=f32, encoding=BYTE_STREAM_SPLIT"

cargo bench -p parquet --bench encoding --all-features -- "decoding: dtype=f64, encoding=BYTE_STREAM_SPLIT"

Are there any user-facing changes?

No.

[etseidl]

run benchmark parquet_round_trip

env:
  BENCH_FILTER: byte_stream_split

[etseidl]

run benchmark arrow_reader

env:
  BENCH_FILTER: byte_stream_split

[adriangbot]

🤖 Arrow criterion benchmark running (GKE) | trigger
Instance: c4a-highmem-16 (12 vCPU / 65 GiB) | Linux bench-c4520212054-286-kpq5v 6.12.68+ #1 SMP Wed Apr 1 02:23:28 UTC 2026 aarch64 GNU/Linux

CPU Details (lscpu)

Architecture:                            aarch64
CPU op-mode(s):                          64-bit
Byte Order:                              Little Endian
CPU(s):                                  16
On-line CPU(s) list:                     0-15
Vendor ID:                               ARM
Model name:                              Neoverse-V2
Model:                                   1
Thread(s) per core:                      1
Core(s) per cluster:                     16
Socket(s):                               -
Cluster(s):                              1
Stepping:                                r0p1
BogoMIPS:                                2000.00
Flags:                                   fp asimd evtstrm aes pmull sha1 sha2 crc32 atomics fphp asimdhp cpuid asimdrdm jscvt fcma lrcpc dcpop sha3 sm3 sm4 asimddp sha512 sve asimdfhm dit uscat ilrcpc flagm sb paca pacg dcpodp sve2 sveaes svepmull svebitperm svesha3 svesm4 flagm2 frint svei8mm svebf16 i8mm bf16 dgh rng bti
L1d cache:                               1 MiB (16 instances)
L1i cache:                               1 MiB (16 instances)
L2 cache:                                32 MiB (16 instances)
L3 cache:                                80 MiB (1 instance)
NUMA node(s):                            1
NUMA node0 CPU(s):                       0-15
Vulnerability Gather data sampling:      Not affected
Vulnerability Indirect target selection: Not affected
Vulnerability Itlb multihit:             Not affected
Vulnerability L1tf:                      Not affected
Vulnerability Mds:                       Not affected
Vulnerability Meltdown:                  Not affected
Vulnerability Mmio stale data:           Not affected
Vulnerability Reg file data sampling:    Not affected
Vulnerability Retbleed:                  Not affected
Vulnerability Spec rstack overflow:      Not affected
Vulnerability Spec store bypass:         Mitigation; Speculative Store Bypass disabled via prctl
Vulnerability Spectre v1:                Mitigation; __user pointer sanitization
Vulnerability Spectre v2:                Mitigation; CSV2, BHB
Vulnerability Srbds:                     Not affected
Vulnerability Tsa:                       Not affected
Vulnerability Tsx async abort:           Not affected
Vulnerability Vmscape:                   Not affected

Comparing parquet-byte-stream-split-cache-locality (3aac60b) to 2f923f7 (merge-base) diff
BENCH_NAME=parquet_round_trip
BENCH_COMMAND=cargo bench --features=arrow,async,test_common,experimental,object_store --bench parquet_round_trip
BENCH_FILTER=byte_stream_split
Results will be posted here when complete

---

File an issue against this benchmark runner

[adriangbot]

🤖 Arrow criterion benchmark running (GKE) | trigger
Instance: c4a-highmem-16 (12 vCPU / 65 GiB) | Linux bench-c4520221803-287-hbjk9 6.12.68+ #1 SMP Wed Apr 1 02:23:28 UTC 2026 aarch64 GNU/Linux

CPU Details (lscpu)

Architecture:                            aarch64
CPU op-mode(s):                          64-bit
Byte Order:                              Little Endian
CPU(s):                                  16
On-line CPU(s) list:                     0-15
Vendor ID:                               ARM
Model name:                              Neoverse-V2
Model:                                   1
Thread(s) per core:                      1
Core(s) per cluster:                     16
Socket(s):                               -
Cluster(s):                              1
Stepping:                                r0p1
BogoMIPS:                                2000.00
Flags:                                   fp asimd evtstrm aes pmull sha1 sha2 crc32 atomics fphp asimdhp cpuid asimdrdm jscvt fcma lrcpc dcpop sha3 sm3 sm4 asimddp sha512 sve asimdfhm dit uscat ilrcpc flagm sb paca pacg dcpodp sve2 sveaes svepmull svebitperm svesha3 svesm4 flagm2 frint svei8mm svebf16 i8mm bf16 dgh rng bti
L1d cache:                               1 MiB (16 instances)
L1i cache:                               1 MiB (16 instances)
L2 cache:                                32 MiB (16 instances)
L3 cache:                                80 MiB (1 instance)
NUMA node(s):                            1
NUMA node0 CPU(s):                       0-15
Vulnerability Gather data sampling:      Not affected
Vulnerability Indirect target selection: Not affected
Vulnerability Itlb multihit:             Not affected
Vulnerability L1tf:                      Not affected
Vulnerability Mds:                       Not affected
Vulnerability Meltdown:                  Not affected
Vulnerability Mmio stale data:           Not affected
Vulnerability Reg file data sampling:    Not affected
Vulnerability Retbleed:                  Not affected
Vulnerability Spec rstack overflow:      Not affected
Vulnerability Spec store bypass:         Mitigation; Speculative Store Bypass disabled via prctl
Vulnerability Spectre v1:                Mitigation; __user pointer sanitization
Vulnerability Spectre v2:                Mitigation; CSV2, BHB
Vulnerability Srbds:                     Not affected
Vulnerability Tsa:                       Not affected
Vulnerability Tsx async abort:           Not affected
Vulnerability Vmscape:                   Not affected

Comparing parquet-byte-stream-split-cache-locality (3aac60b) to 2f923f7 (merge-base) diff
BENCH_NAME=arrow_reader
BENCH_COMMAND=cargo bench --features=arrow,async,test_common,experimental,object_store --bench arrow_reader
BENCH_FILTER=byte_stream_split
Results will be posted here when complete

---

File an issue against this benchmark runner

[adriangbot]

🤖 Arrow criterion benchmark completed (GKE) | trigger

Instance: c4a-highmem-16 (12 vCPU / 65 GiB)

CPU Details (lscpu)

Architecture:                            aarch64
CPU op-mode(s):                          64-bit
Byte Order:                              Little Endian
CPU(s):                                  16
On-line CPU(s) list:                     0-15
Vendor ID:                               ARM
Model name:                              Neoverse-V2
Model:                                   1
Thread(s) per core:                      1
Core(s) per cluster:                     16
Socket(s):                               -
Cluster(s):                              1
Stepping:                                r0p1
BogoMIPS:                                2000.00
Flags:                                   fp asimd evtstrm aes pmull sha1 sha2 crc32 atomics fphp asimdhp cpuid asimdrdm jscvt fcma lrcpc dcpop sha3 sm3 sm4 asimddp sha512 sve asimdfhm dit uscat ilrcpc flagm sb paca pacg dcpodp sve2 sveaes svepmull svebitperm svesha3 svesm4 flagm2 frint svei8mm svebf16 i8mm bf16 dgh rng bti
L1d cache:                               1 MiB (16 instances)
L1i cache:                               1 MiB (16 instances)
L2 cache:                                32 MiB (16 instances)
L3 cache:                                80 MiB (1 instance)
NUMA node(s):                            1
NUMA node0 CPU(s):                       0-15
Vulnerability Gather data sampling:      Not affected
Vulnerability Indirect target selection: Not affected
Vulnerability Itlb multihit:             Not affected
Vulnerability L1tf:                      Not affected
Vulnerability Mds:                       Not affected
Vulnerability Meltdown:                  Not affected
Vulnerability Mmio stale data:           Not affected
Vulnerability Reg file data sampling:    Not affected
Vulnerability Retbleed:                  Not affected
Vulnerability Spec rstack overflow:      Not affected
Vulnerability Spec store bypass:         Mitigation; Speculative Store Bypass disabled via prctl
Vulnerability Spectre v1:                Mitigation; __user pointer sanitization
Vulnerability Spectre v2:                Mitigation; CSV2, BHB
Vulnerability Srbds:                     Not affected
Vulnerability Tsa:                       Not affected
Vulnerability Tsx async abort:           Not affected
Vulnerability Vmscape:                   Not affected

Details

group                                main                                   parquet-byte-stream-split-cache-locality
-----                                ----                                   ----------------------------------------
read Fixed(16) byte_stream_split     1.01      5.5±0.03ms        ? ?/sec    1.00      5.5±0.07ms        ? ?/sec
read Fixed(2) byte_stream_split      1.00    929.8±3.83µs        ? ?/sec    1.00    932.2±2.33µs        ? ?/sec
read f32 byte_stream_split           1.00   1014.9±2.41µs        ? ?/sec    2.16      2.2±0.01ms        ? ?/sec
read f64 byte_stream_split           1.00  1563.2±38.91µs        ? ?/sec    2.63      4.1±0.20ms        ? ?/sec
read int32 byte_stream_split         1.00   1087.3±6.98µs        ? ?/sec    2.07      2.2±0.04ms        ? ?/sec
read int64 byte_stream_split         1.00  1581.8±19.86µs        ? ?/sec    2.51      4.0±0.00ms        ? ?/sec
write Fixed(16) byte_stream_split    1.00     27.7±0.17ms        ? ?/sec    1.00     27.6±0.12ms        ? ?/sec
write Fixed(2) byte_stream_split     1.00     21.7±0.08ms        ? ?/sec    1.00     21.6±0.09ms        ? ?/sec
write f32 byte_stream_split          1.00      5.0±0.03ms        ? ?/sec    1.01      5.0±0.04ms        ? ?/sec
write f64 byte_stream_split          1.00      6.8±0.04ms        ? ?/sec    1.00      6.8±0.04ms        ? ?/sec
write int32 byte_stream_split        1.01      5.4±0.08ms        ? ?/sec    1.00      5.4±0.04ms        ? ?/sec
write int64 byte_stream_split        1.01      6.9±0.06ms        ? ?/sec    1.00      6.9±0.03ms        ? ?/sec

Resource Usage

base (merge-base)

Metric	Value
Wall time	135.0s
Peak memory	6.1 GiB
Avg memory	5.3 GiB
CPU user	128.1s
CPU sys	3.2s
Peak spill	0 B

branch

Metric	Value
Wall time	125.0s
Peak memory	6.1 GiB
Avg memory	5.2 GiB
CPU user	121.1s
CPU sys	2.5s
Peak spill	0 B

---

File an issue against this benchmark runner

[adriangbot]

🤖 Arrow criterion benchmark completed (GKE) | trigger

Instance: c4a-highmem-16 (12 vCPU / 65 GiB)

CPU Details (lscpu)

Architecture:                            aarch64
CPU op-mode(s):                          64-bit
Byte Order:                              Little Endian
CPU(s):                                  16
On-line CPU(s) list:                     0-15
Vendor ID:                               ARM
Model name:                              Neoverse-V2
Model:                                   1
Thread(s) per core:                      1
Core(s) per cluster:                     16
Socket(s):                               -
Cluster(s):                              1
Stepping:                                r0p1
BogoMIPS:                                2000.00
Flags:                                   fp asimd evtstrm aes pmull sha1 sha2 crc32 atomics fphp asimdhp cpuid asimdrdm jscvt fcma lrcpc dcpop sha3 sm3 sm4 asimddp sha512 sve asimdfhm dit uscat ilrcpc flagm sb paca pacg dcpodp sve2 sveaes svepmull svebitperm svesha3 svesm4 flagm2 frint svei8mm svebf16 i8mm bf16 dgh rng bti
L1d cache:                               1 MiB (16 instances)
L1i cache:                               1 MiB (16 instances)
L2 cache:                                32 MiB (16 instances)
L3 cache:                                80 MiB (1 instance)
NUMA node(s):                            1
NUMA node0 CPU(s):                       0-15
Vulnerability Gather data sampling:      Not affected
Vulnerability Indirect target selection: Not affected
Vulnerability Itlb multihit:             Not affected
Vulnerability L1tf:                      Not affected
Vulnerability Mds:                       Not affected
Vulnerability Meltdown:                  Not affected
Vulnerability Mmio stale data:           Not affected
Vulnerability Reg file data sampling:    Not affected
Vulnerability Retbleed:                  Not affected
Vulnerability Spec rstack overflow:      Not affected
Vulnerability Spec store bypass:         Mitigation; Speculative Store Bypass disabled via prctl
Vulnerability Spectre v1:                Mitigation; __user pointer sanitization
Vulnerability Spectre v2:                Mitigation; CSV2, BHB
Vulnerability Srbds:                     Not affected
Vulnerability Tsa:                       Not affected
Vulnerability Tsx async abort:           Not affected
Vulnerability Vmscape:                   Not affected

Details

group                                                                                                      main                                   parquet-byte-stream-split-cache-locality
-----                                                                                                      ----                                   ----------------------------------------
arrow_array_reader/FIXED_LEN_BYTE_ARRAY/Decimal128Array/byte_stream_split encoded, mandatory, no NULLs     1.11  1076.1±14.88µs        ? ?/sec    1.00   968.6±12.93µs        ? ?/sec
arrow_array_reader/FIXED_LEN_BYTE_ARRAY/Decimal128Array/byte_stream_split encoded, optional, half NULLs    1.00    670.1±0.50µs        ? ?/sec    1.00    669.5±0.63µs        ? ?/sec
arrow_array_reader/FIXED_LEN_BYTE_ARRAY/Decimal128Array/byte_stream_split encoded, optional, no NULLs      1.00    971.7±0.66µs        ? ?/sec    1.03  1005.6±21.67µs        ? ?/sec
arrow_array_reader/FIXED_LEN_BYTE_ARRAY/Float16Array/byte_stream_split encoded, mandatory, no NULLs        1.00    128.4±0.21µs        ? ?/sec    1.01    129.2±0.96µs        ? ?/sec
arrow_array_reader/FIXED_LEN_BYTE_ARRAY/Float16Array/byte_stream_split encoded, optional, half NULLs       1.00    183.2±0.18µs        ? ?/sec    1.01    185.4±1.65µs        ? ?/sec
arrow_array_reader/FIXED_LEN_BYTE_ARRAY/Float16Array/byte_stream_split encoded, optional, no NULLs         1.00    132.3±0.27µs        ? ?/sec    1.07    141.5±3.17µs        ? ?/sec
arrow_array_reader/FixedLenByteArray(16)/byte_stream_split encoded, mandatory, no NULLs                    1.00    801.5±0.42µs        ? ?/sec    1.00    802.0±0.63µs        ? ?/sec
arrow_array_reader/FixedLenByteArray(16)/byte_stream_split encoded, optional, half NULLs                   1.00    508.6±0.39µs        ? ?/sec    1.00    507.9±0.55µs        ? ?/sec
arrow_array_reader/FixedLenByteArray(16)/byte_stream_split encoded, optional, no NULLs                     1.00    807.3±0.73µs        ? ?/sec    1.05   844.1±62.76µs        ? ?/sec
arrow_array_reader/FixedLenByteArray(2)/byte_stream_split encoded, mandatory, no NULLs                     1.03     75.8±0.21µs        ? ?/sec    1.00     73.4±0.34µs        ? ?/sec
arrow_array_reader/FixedLenByteArray(2)/byte_stream_split encoded, optional, half NULLs                    1.01    128.9±0.19µs        ? ?/sec    1.00    128.0±0.22µs        ? ?/sec
arrow_array_reader/FixedLenByteArray(2)/byte_stream_split encoded, optional, no NULLs                      1.01     77.0±0.42µs        ? ?/sec    1.00     76.5±0.47µs        ? ?/sec
arrow_array_reader/FixedLenByteArray(4)/byte_stream_split encoded, mandatory, no NULLs                     1.02    150.1±0.43µs        ? ?/sec    1.00    147.6±0.76µs        ? ?/sec
arrow_array_reader/FixedLenByteArray(4)/byte_stream_split encoded, optional, half NULLs                    1.00    198.3±0.23µs        ? ?/sec    1.01    199.6±0.32µs        ? ?/sec
arrow_array_reader/FixedLenByteArray(4)/byte_stream_split encoded, optional, no NULLs                      1.03    153.2±0.87µs        ? ?/sec    1.00    149.1±0.39µs        ? ?/sec
arrow_array_reader/FixedLenByteArray(8)/byte_stream_split encoded, mandatory, no NULLs                     1.02    303.1±0.83µs        ? ?/sec    1.00    295.8±1.30µs        ? ?/sec
arrow_array_reader/FixedLenByteArray(8)/byte_stream_split encoded, optional, half NULLs                    1.00    265.6±0.39µs        ? ?/sec    1.00    265.3±0.72µs        ? ?/sec
arrow_array_reader/FixedLenByteArray(8)/byte_stream_split encoded, optional, no NULLs                      1.00    304.4±1.07µs        ? ?/sec    1.00    302.9±1.08µs        ? ?/sec
arrow_array_reader/INT32/Decimal128Array/byte_stream_split encoded, mandatory, no NULLs                    1.00     54.7±0.06µs        ? ?/sec    2.67    145.9±0.25µs        ? ?/sec
arrow_array_reader/INT32/Decimal128Array/byte_stream_split encoded, optional, half NULLs                   1.00    119.1±0.35µs        ? ?/sec    1.39    165.6±0.10µs        ? ?/sec
arrow_array_reader/INT32/Decimal128Array/byte_stream_split encoded, optional, no NULLs                     1.00     57.9±0.10µs        ? ?/sec    2.58    149.3±0.11µs        ? ?/sec
arrow_array_reader/INT64/Decimal128Array/byte_stream_split encoded, mandatory, no NULLs                    1.00     82.7±0.09µs        ? ?/sec    3.55    293.8±3.05µs        ? ?/sec
arrow_array_reader/INT64/Decimal128Array/byte_stream_split encoded, optional, half NULLs                   1.00    127.9±0.18µs        ? ?/sec    1.82    232.2±1.53µs        ? ?/sec
arrow_array_reader/INT64/Decimal128Array/byte_stream_split encoded, optional, no NULLs                     1.00     87.6±0.08µs        ? ?/sec    3.40    297.7±3.03µs        ? ?/sec
arrow_array_reader/Int16Array/byte_stream_split encoded, mandatory, no NULLs                               1.00     21.0±0.03µs        ? ?/sec    5.39    112.9±0.10µs        ? ?/sec
arrow_array_reader/Int16Array/byte_stream_split encoded, optional, half NULLs                              1.00     86.6±0.10µs        ? ?/sec    1.53    132.3±0.07µs        ? ?/sec
arrow_array_reader/Int16Array/byte_stream_split encoded, optional, no NULLs                                1.00     24.9±0.04µs        ? ?/sec    4.66    116.2±0.10µs        ? ?/sec
arrow_array_reader/Int32Array/byte_stream_split encoded, mandatory, no NULLs                               1.00     17.3±0.07µs        ? ?/sec    6.35    110.0±0.47µs        ? ?/sec
arrow_array_reader/Int32Array/byte_stream_split encoded, optional, half NULLs                              1.00     82.8±0.20µs        ? ?/sec    1.55    128.5±0.10µs        ? ?/sec
arrow_array_reader/Int32Array/byte_stream_split encoded, optional, no NULLs                                1.00     20.9±0.11µs        ? ?/sec    5.39    112.9±0.37µs        ? ?/sec
arrow_array_reader/Int64Array/byte_stream_split encoded, mandatory, no NULLs                               1.00     45.6±0.04µs        ? ?/sec    5.65    257.7±3.26µs        ? ?/sec
arrow_array_reader/Int64Array/byte_stream_split encoded, optional, half NULLs                              1.00     91.5±0.12µs        ? ?/sec    2.16    197.8±1.59µs        ? ?/sec
arrow_array_reader/Int64Array/byte_stream_split encoded, optional, no NULLs                                1.00     49.9±0.04µs        ? ?/sec    5.25    261.8±3.26µs        ? ?/sec
arrow_array_reader/Int8Array/byte_stream_split encoded, mandatory, no NULLs                                1.00     20.7±0.08µs        ? ?/sec    5.43    112.6±0.10µs        ? ?/sec
arrow_array_reader/Int8Array/byte_stream_split encoded, optional, half NULLs                               1.00     85.7±0.07µs        ? ?/sec    1.54    131.7±0.09µs        ? ?/sec
arrow_array_reader/Int8Array/byte_stream_split encoded, optional, no NULLs                                 1.00     24.7±0.07µs        ? ?/sec    4.70    115.9±0.08µs        ? ?/sec
arrow_array_reader/UInt16Array/byte_stream_split encoded, mandatory, no NULLs                              1.00     21.0±0.10µs        ? ?/sec    5.39    113.0±0.13µs        ? ?/sec
arrow_array_reader/UInt16Array/byte_stream_split encoded, optional, half NULLs                             1.00     86.6±0.08µs        ? ?/sec    1.53    132.6±0.12µs        ? ?/sec
arrow_array_reader/UInt16Array/byte_stream_split encoded, optional, no NULLs                               1.00     24.9±0.06µs        ? ?/sec    4.67    116.5±0.11µs        ? ?/sec
arrow_array_reader/UInt32Array/byte_stream_split encoded, mandatory, no NULLs                              1.00     17.5±0.01µs        ? ?/sec    6.25    109.7±0.09µs        ? ?/sec
arrow_array_reader/UInt32Array/byte_stream_split encoded, optional, half NULLs                             1.00     82.6±0.07µs        ? ?/sec    1.57    129.6±0.09µs        ? ?/sec
arrow_array_reader/UInt32Array/byte_stream_split encoded, optional, no NULLs                               1.00     21.1±0.03µs        ? ?/sec    5.36    113.1±0.08µs        ? ?/sec
arrow_array_reader/UInt64Array/byte_stream_split encoded, mandatory, no NULLs                              1.00     45.7±0.03µs        ? ?/sec    5.63    257.7±3.22µs        ? ?/sec
arrow_array_reader/UInt64Array/byte_stream_split encoded, optional, half NULLs                             1.00     91.8±0.13µs        ? ?/sec    2.13    195.4±1.51µs        ? ?/sec
arrow_array_reader/UInt64Array/byte_stream_split encoded, optional, no NULLs                               1.00     50.1±0.11µs        ? ?/sec    5.23    261.6±3.25µs        ? ?/sec
arrow_array_reader/UInt8Array/byte_stream_split encoded, mandatory, no NULLs                               1.00     20.9±0.05µs        ? ?/sec    5.38    112.7±0.09µs        ? ?/sec
arrow_array_reader/UInt8Array/byte_stream_split encoded, optional, half NULLs                              1.00     85.5±0.19µs        ? ?/sec    1.55    132.2±0.07µs        ? ?/sec
arrow_array_reader/UInt8Array/byte_stream_split encoded, optional, no NULLs                                1.00     24.4±0.06µs        ? ?/sec    4.74    115.9±0.11µs        ? ?/sec

Resource Usage

base (merge-base)

Metric	Value
Wall time	500.1s
Peak memory	4.2 GiB
Avg memory	4.2 GiB
CPU user	497.7s
CPU sys	0.8s
Peak spill	0 B

branch

Metric	Value
Wall time	485.1s
Peak memory	4.2 GiB
Avg memory	4.2 GiB
CPU user	481.1s
CPU sys	0.2s
Peak spill	0 B

---

File an issue against this benchmark runner