Test performance of BatchCoalescer::push_batch_with_filter

[alamb]

This PR is to test the end to end query performance for @ClSlaid 's PR in arrow:

• arrow-select: Optimize BatchCoalescer::push_batch_with_filter for low selectivity filters and inlined Utf8View/BinaryView arrow-rs#9755

[alamb]

Run benchmarks

[adriangbot]

🤖 Benchmark running (GKE) | trigger
Instance: c4a-highmem-16 (12 vCPU / 65 GiB) | Linux bench-c4662575972-516-47s44 6.12.68+ #1 SMP Sat May 2 07:49:07 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 alamb/test_coalesce (5e19ac2) to 8995ce6 (merge-base) diff using: tpch
Results will be posted here when complete

---

File an issue against this benchmark runner

[adriangbot]

🤖 Benchmark running (GKE) | trigger
Instance: c4a-highmem-16 (12 vCPU / 65 GiB) | Linux bench-c4662575972-515-nq7xr 6.12.68+ #1 SMP Sat May 2 07:49:07 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 alamb/test_coalesce (5e19ac2) to 8995ce6 (merge-base) diff using: tpcds
Results will be posted here when complete

---

File an issue against this benchmark runner

[adriangbot]

🤖 Benchmark running (GKE) | trigger
Instance: c4a-highmem-16 (12 vCPU / 65 GiB) | Linux bench-c4662575972-514-7jndj 6.12.68+ #1 SMP Sat May 2 07:49:07 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 alamb/test_coalesce (5e19ac2) to 8995ce6 (merge-base) diff using: clickbench_partitioned
Results will be posted here when complete

---

File an issue against this benchmark runner

[adriangbot]

🤖 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

Comparing HEAD and alamb_test_coalesce
--------------------
Benchmark tpch_sf1.json
--------------------
┏━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━┓
┃ Query     ┃                           HEAD ┃              alamb_test_coalesce ┃       Change ┃
┡━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━┩
│ QQuery 1  │ 38.62 / 39.39 ±1.03 / 41.41 ms │ 37.86 / 86.82 ±96.92 / 280.65 ms │ 2.20x slower │
│ QQuery 2  │ 18.89 / 19.10 ±0.26 / 19.60 ms │   18.48 / 19.05 ±0.63 / 20.26 ms │    no change │
│ QQuery 3  │ 31.05 / 32.49 ±1.49 / 34.50 ms │   30.30 / 32.81 ±2.28 / 36.65 ms │    no change │
│ QQuery 4  │ 17.42 / 17.89 ±0.72 / 19.31 ms │   16.92 / 17.39 ±0.67 / 18.69 ms │    no change │
│ QQuery 5  │ 38.60 / 40.40 ±0.94 / 41.35 ms │   37.56 / 39.07 ±1.71 / 41.36 ms │    no change │
│ QQuery 6  │ 16.15 / 16.30 ±0.09 / 16.42 ms │   15.78 / 17.14 ±1.20 / 18.77 ms │ 1.05x slower │
│ QQuery 7  │ 42.80 / 44.74 ±1.80 / 48.09 ms │   43.94 / 45.69 ±1.39 / 47.38 ms │    no change │
│ QQuery 8  │ 42.69 / 43.28 ±0.55 / 44.21 ms │   42.52 / 42.81 ±0.47 / 43.74 ms │    no change │
│ QQuery 9  │ 48.95 / 49.49 ±0.90 / 51.28 ms │   48.61 / 50.03 ±0.93 / 50.88 ms │    no change │
│ QQuery 10 │ 41.97 / 42.22 ±0.25 / 42.62 ms │   41.55 / 42.97 ±1.31 / 44.65 ms │    no change │
│ QQuery 11 │ 13.20 / 13.33 ±0.12 / 13.54 ms │   12.91 / 14.09 ±1.59 / 17.24 ms │ 1.06x slower │
│ QQuery 12 │ 23.85 / 24.67 ±0.52 / 25.42 ms │   23.70 / 24.46 ±0.51 / 25.20 ms │    no change │
│ QQuery 13 │ 31.85 / 33.03 ±0.67 / 33.84 ms │   31.80 / 33.70 ±1.65 / 35.96 ms │    no change │
│ QQuery 14 │ 23.72 / 23.88 ±0.13 / 24.05 ms │   23.48 / 23.84 ±0.37 / 24.50 ms │    no change │
│ QQuery 15 │ 31.27 / 31.49 ±0.14 / 31.65 ms │   30.73 / 30.84 ±0.09 / 30.96 ms │    no change │
│ QQuery 16 │ 13.84 / 14.03 ±0.11 / 14.18 ms │   14.01 / 14.36 ±0.24 / 14.65 ms │    no change │
│ QQuery 17 │ 72.51 / 73.92 ±1.52 / 76.73 ms │   72.45 / 73.74 ±0.80 / 74.56 ms │    no change │
│ QQuery 18 │ 57.06 / 60.58 ±3.57 / 67.30 ms │   56.10 / 59.16 ±1.79 / 61.17 ms │    no change │
│ QQuery 19 │ 32.95 / 33.29 ±0.40 / 33.96 ms │   32.35 / 32.81 ±0.65 / 34.08 ms │    no change │
│ QQuery 20 │ 31.98 / 32.28 ±0.26 / 32.65 ms │   31.52 / 32.24 ±0.58 / 33.18 ms │    no change │
│ QQuery 21 │ 55.05 / 57.32 ±1.85 / 60.47 ms │   54.42 / 57.38 ±2.08 / 60.04 ms │    no change │
│ QQuery 22 │ 13.80 / 14.15 ±0.41 / 14.91 ms │   13.93 / 15.03 ±1.96 / 18.96 ms │ 1.06x slower │
└───────────┴────────────────────────────────┴──────────────────────────────────┴──────────────┘
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━┓
┃ Benchmark Summary                  ┃          ┃
┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━┩
│ Total Time (HEAD)                  │ 757.27ms │
│ Total Time (alamb_test_coalesce)   │ 805.43ms │
│ Average Time (HEAD)                │  34.42ms │
│ Average Time (alamb_test_coalesce) │  36.61ms │
│ Queries Faster                     │        0 │
│ Queries Slower                     │        4 │
│ Queries with No Change             │       18 │
│ Queries with Failure               │        0 │
└────────────────────────────────────┴──────────┘

Resource Usage

tpch — base (merge-base)

Metric	Value
Wall time	5.0s
Peak memory	1.1 GiB
Avg memory	508.6 MiB
CPU user	22.1s
CPU sys	1.8s
Peak spill	0 B

tpch — branch

Metric	Value
Wall time	5.0s
Peak memory	1.3 GiB
Avg memory	699.3 MiB
CPU user	23.9s
CPU sys	1.9s
Peak spill	0 B

---

File an issue against this benchmark runner

[adriangbot]

🤖 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

Comparing HEAD and alamb_test_coalesce
--------------------
Benchmark tpcds_sf1.json
--------------------
┏━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓
┃ Query     ┃                                  HEAD ┃                   alamb_test_coalesce ┃        Change ┃
┡━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩
│ QQuery 1  │           6.21 / 6.75 ±0.92 / 8.59 ms │           6.04 / 6.49 ±0.82 / 8.13 ms │     no change │
│ QQuery 2  │        83.13 / 83.53 ±0.30 / 83.97 ms │        80.92 / 81.19 ±0.21 / 81.44 ms │     no change │
│ QQuery 3  │        29.57 / 30.24 ±0.40 / 30.78 ms │        29.09 / 29.32 ±0.22 / 29.59 ms │     no change │
│ QQuery 4  │     521.69 / 530.57 ±4.76 / 535.01 ms │     507.41 / 509.78 ±2.59 / 514.61 ms │     no change │
│ QQuery 5  │        53.40 / 53.74 ±0.49 / 54.70 ms │        51.56 / 52.02 ±0.24 / 52.25 ms │     no change │
│ QQuery 6  │        38.22 / 38.57 ±0.42 / 39.33 ms │        37.33 / 37.56 ±0.16 / 37.78 ms │     no change │
│ QQuery 7  │      97.92 / 101.03 ±3.49 / 107.74 ms │        96.28 / 96.77 ±0.29 / 97.15 ms │     no change │
│ QQuery 8  │        38.04 / 38.40 ±0.33 / 38.93 ms │        37.45 / 37.81 ±0.39 / 38.53 ms │     no change │
│ QQuery 9  │        54.75 / 56.24 ±1.11 / 57.77 ms │        53.49 / 56.68 ±3.26 / 62.66 ms │     no change │
│ QQuery 10 │        63.85 / 64.21 ±0.36 / 64.82 ms │        62.91 / 63.17 ±0.24 / 63.58 ms │     no change │
│ QQuery 11 │     318.16 / 327.77 ±8.55 / 342.86 ms │     323.45 / 327.01 ±3.36 / 332.16 ms │     no change │
│ QQuery 12 │        29.65 / 30.00 ±0.30 / 30.41 ms │        29.01 / 29.41 ±0.36 / 30.07 ms │     no change │
│ QQuery 13 │     121.11 / 122.37 ±0.97 / 123.80 ms │     119.28 / 120.52 ±1.94 / 124.38 ms │     no change │
│ QQuery 14 │     420.13 / 426.71 ±6.83 / 439.34 ms │     413.75 / 417.62 ±2.80 / 421.51 ms │     no change │
│ QQuery 15 │        61.07 / 62.77 ±1.26 / 64.55 ms │        59.81 / 60.84 ±0.86 / 62.03 ms │     no change │
│ QQuery 16 │           7.12 / 7.33 ±0.23 / 7.78 ms │           6.79 / 7.01 ±0.21 / 7.39 ms │     no change │
│ QQuery 17 │        82.19 / 84.76 ±2.15 / 88.14 ms │        81.09 / 81.97 ±1.46 / 84.88 ms │     no change │
│ QQuery 18 │     127.74 / 128.83 ±0.90 / 130.44 ms │     127.37 / 127.86 ±0.43 / 128.52 ms │     no change │
│ QQuery 19 │        43.41 / 44.22 ±1.09 / 46.37 ms │        42.14 / 42.67 ±0.35 / 43.23 ms │     no change │
│ QQuery 20 │        37.03 / 37.89 ±0.48 / 38.45 ms │        36.62 / 37.91 ±1.38 / 40.52 ms │     no change │
│ QQuery 21 │        17.70 / 18.26 ±0.34 / 18.77 ms │        17.44 / 18.72 ±2.28 / 23.27 ms │     no change │
│ QQuery 22 │        63.92 / 64.77 ±0.61 / 65.83 ms │        64.33 / 64.79 ±0.47 / 65.71 ms │     no change │
│ QQuery 23 │     328.33 / 334.34 ±5.78 / 344.77 ms │     323.11 / 327.98 ±3.22 / 332.42 ms │     no change │
│ QQuery 24 │     229.49 / 234.30 ±5.10 / 243.06 ms │     226.03 / 229.83 ±3.23 / 235.37 ms │     no change │
│ QQuery 25 │     113.10 / 114.99 ±1.55 / 117.21 ms │     112.29 / 115.59 ±4.35 / 124.00 ms │     no change │
│ QQuery 26 │        59.40 / 60.20 ±1.03 / 62.20 ms │        59.08 / 60.44 ±1.49 / 63.29 ms │     no change │
│ QQuery 27 │           7.04 / 7.14 ±0.12 / 7.37 ms │           6.73 / 7.02 ±0.16 / 7.24 ms │     no change │
│ QQuery 28 │        57.38 / 61.53 ±2.15 / 63.27 ms │        57.67 / 60.53 ±1.56 / 61.95 ms │     no change │
│ QQuery 29 │     100.19 / 103.25 ±2.76 / 108.21 ms │      99.71 / 101.52 ±1.86 / 105.04 ms │     no change │
│ QQuery 30 │        33.76 / 34.26 ±0.39 / 34.97 ms │        33.66 / 35.26 ±1.51 / 37.97 ms │     no change │
│ QQuery 31 │     114.03 / 115.45 ±2.02 / 119.44 ms │     112.95 / 113.31 ±0.29 / 113.72 ms │     no change │
│ QQuery 32 │        20.84 / 21.60 ±0.96 / 23.50 ms │        21.24 / 22.72 ±2.59 / 27.89 ms │  1.05x slower │
│ QQuery 33 │        39.52 / 39.79 ±0.19 / 40.10 ms │        38.52 / 39.40 ±0.51 / 39.91 ms │     no change │
│ QQuery 34 │         9.93 / 10.20 ±0.36 / 10.87 ms │         9.69 / 10.35 ±0.52 / 11.03 ms │     no change │
│ QQuery 35 │        72.77 / 73.49 ±0.43 / 73.87 ms │        72.12 / 74.07 ±1.28 / 76.04 ms │     no change │
│ QQuery 36 │           6.17 / 6.34 ±0.17 / 6.64 ms │           6.07 / 6.20 ±0.16 / 6.52 ms │     no change │
│ QQuery 37 │           6.96 / 7.11 ±0.10 / 7.26 ms │           7.02 / 7.11 ±0.06 / 7.17 ms │     no change │
│ QQuery 38 │        64.63 / 65.63 ±1.15 / 67.86 ms │        64.83 / 65.79 ±1.04 / 67.81 ms │     no change │
│ QQuery 39 │     467.08 / 473.06 ±5.89 / 483.62 ms │     473.55 / 475.72 ±2.40 / 479.56 ms │     no change │
│ QQuery 40 │        23.73 / 24.06 ±0.26 / 24.52 ms │        23.01 / 23.31 ±0.27 / 23.78 ms │     no change │
│ QQuery 41 │        12.07 / 12.33 ±0.29 / 12.70 ms │        11.45 / 11.60 ±0.13 / 11.78 ms │ +1.06x faster │
│ QQuery 42 │        24.83 / 24.98 ±0.15 / 25.25 ms │        23.89 / 24.18 ±0.17 / 24.37 ms │     no change │
│ QQuery 43 │           5.25 / 5.46 ±0.20 / 5.83 ms │           4.87 / 4.97 ±0.14 / 5.24 ms │ +1.10x faster │
│ QQuery 44 │        10.28 / 10.41 ±0.19 / 10.79 ms │           9.11 / 9.30 ±0.16 / 9.52 ms │ +1.12x faster │
│ QQuery 45 │        40.78 / 43.01 ±2.84 / 48.56 ms │        40.71 / 42.29 ±2.66 / 47.59 ms │     no change │
│ QQuery 46 │        12.88 / 13.90 ±1.63 / 17.16 ms │        11.96 / 12.40 ±0.36 / 12.82 ms │ +1.12x faster │
│ QQuery 47 │     239.97 / 248.01 ±4.78 / 252.89 ms │     234.79 / 242.41 ±4.87 / 249.00 ms │     no change │
│ QQuery 48 │        97.73 / 98.10 ±0.29 / 98.50 ms │        96.21 / 97.05 ±0.54 / 97.64 ms │     no change │
│ QQuery 49 │        80.01 / 83.47 ±3.17 / 87.35 ms │        77.69 / 78.30 ±0.45 / 78.80 ms │ +1.07x faster │
│ QQuery 50 │        60.71 / 61.07 ±0.29 / 61.44 ms │        59.72 / 61.72 ±2.83 / 67.28 ms │     no change │
│ QQuery 51 │       94.36 / 96.30 ±2.50 / 101.08 ms │        93.64 / 94.59 ±1.52 / 97.61 ms │     no change │
│ QQuery 52 │        24.54 / 25.93 ±1.71 / 29.25 ms │        24.84 / 25.45 ±0.52 / 26.37 ms │     no change │
│ QQuery 53 │        30.21 / 30.54 ±0.40 / 31.28 ms │        29.59 / 30.33 ±0.49 / 31.06 ms │     no change │
│ QQuery 54 │        56.35 / 56.98 ±0.37 / 57.44 ms │        56.71 / 58.15 ±1.92 / 61.96 ms │     no change │
│ QQuery 55 │        23.97 / 24.29 ±0.25 / 24.73 ms │        23.92 / 24.41 ±0.36 / 24.93 ms │     no change │
│ QQuery 56 │        40.43 / 40.78 ±0.32 / 41.39 ms │        39.47 / 39.77 ±0.37 / 40.49 ms │     no change │
│ QQuery 57 │     179.44 / 182.44 ±2.74 / 185.77 ms │     179.55 / 183.26 ±2.74 / 187.64 ms │     no change │
│ QQuery 58 │     118.55 / 120.19 ±2.01 / 123.83 ms │     117.28 / 119.39 ±1.54 / 121.30 ms │     no change │
│ QQuery 59 │     119.95 / 122.59 ±3.11 / 128.61 ms │     117.60 / 119.06 ±1.01 / 120.21 ms │     no change │
│ QQuery 60 │        41.68 / 42.11 ±0.27 / 42.50 ms │        40.09 / 41.14 ±1.17 / 43.40 ms │     no change │
│ QQuery 61 │        13.90 / 14.03 ±0.09 / 14.16 ms │        13.34 / 13.48 ±0.11 / 13.67 ms │     no change │
│ QQuery 62 │        47.36 / 47.71 ±0.21 / 48.00 ms │        46.88 / 47.76 ±1.19 / 50.05 ms │     no change │
│ QQuery 63 │        30.08 / 30.58 ±0.30 / 31.04 ms │        29.94 / 30.38 ±0.40 / 31.10 ms │     no change │
│ QQuery 64 │     404.81 / 409.21 ±3.73 / 414.62 ms │     399.76 / 402.96 ±3.60 / 409.66 ms │     no change │
│ QQuery 65 │     149.83 / 159.54 ±8.70 / 172.61 ms │     152.89 / 156.06 ±2.73 / 161.02 ms │     no change │
│ QQuery 66 │        81.45 / 82.14 ±0.60 / 83.22 ms │        81.22 / 81.60 ±0.34 / 82.09 ms │     no change │
│ QQuery 67 │     258.92 / 269.73 ±6.45 / 276.31 ms │     258.18 / 262.32 ±4.14 / 269.77 ms │     no change │
│ QQuery 68 │        12.39 / 12.63 ±0.18 / 12.89 ms │        12.22 / 12.38 ±0.13 / 12.62 ms │     no change │
│ QQuery 69 │        57.85 / 58.56 ±0.39 / 59.04 ms │        57.43 / 58.41 ±0.70 / 59.61 ms │     no change │
│ QQuery 70 │    107.62 / 116.53 ±12.48 / 140.01 ms │     106.02 / 107.59 ±1.34 / 109.51 ms │ +1.08x faster │
│ QQuery 71 │        36.63 / 37.00 ±0.36 / 37.63 ms │        36.07 / 38.42 ±3.61 / 45.60 ms │     no change │
│ QQuery 72 │ 2222.21 / 2272.18 ±33.27 / 2326.65 ms │ 2094.30 / 2173.46 ±46.55 / 2230.88 ms │     no change │
│ QQuery 73 │          9.68 / 9.92 ±0.19 / 10.16 ms │          9.49 / 9.81 ±0.23 / 10.18 ms │     no change │
│ QQuery 74 │     183.73 / 191.23 ±5.30 / 198.33 ms │     180.84 / 183.82 ±3.61 / 190.93 ms │     no change │
│ QQuery 75 │     150.77 / 153.49 ±1.53 / 155.09 ms │     149.20 / 150.27 ±0.82 / 151.53 ms │     no change │
│ QQuery 76 │        36.46 / 36.91 ±0.50 / 37.83 ms │        35.62 / 35.91 ±0.20 / 36.20 ms │     no change │
│ QQuery 77 │        63.45 / 64.38 ±0.81 / 65.63 ms │        61.45 / 64.22 ±3.56 / 71.22 ms │     no change │
│ QQuery 78 │     191.44 / 196.97 ±4.38 / 204.49 ms │     188.43 / 194.55 ±5.48 / 204.09 ms │     no change │
│ QQuery 79 │        68.56 / 71.51 ±4.32 / 80.00 ms │        67.62 / 69.17 ±1.66 / 72.12 ms │     no change │
│ QQuery 80 │     102.84 / 108.00 ±4.73 / 115.19 ms │     101.45 / 102.94 ±1.34 / 105.39 ms │     no change │
│ QQuery 81 │        26.76 / 26.98 ±0.21 / 27.25 ms │        26.12 / 26.36 ±0.19 / 26.65 ms │     no change │
│ QQuery 82 │        17.20 / 18.53 ±2.24 / 23.00 ms │        16.75 / 16.89 ±0.15 / 17.14 ms │ +1.10x faster │
│ QQuery 83 │        41.33 / 41.84 ±0.37 / 42.19 ms │        40.65 / 42.73 ±2.35 / 47.02 ms │     no change │
│ QQuery 84 │        35.18 / 35.47 ±0.23 / 35.75 ms │        34.76 / 35.37 ±0.32 / 35.66 ms │     no change │
│ QQuery 85 │     110.04 / 111.54 ±1.94 / 115.25 ms │     108.11 / 110.83 ±3.01 / 116.56 ms │     no change │
│ QQuery 86 │        25.98 / 28.85 ±4.58 / 37.95 ms │        25.68 / 26.25 ±0.55 / 27.05 ms │ +1.10x faster │
│ QQuery 87 │        66.73 / 68.02 ±1.09 / 69.37 ms │        66.09 / 66.84 ±0.40 / 67.22 ms │     no change │
│ QQuery 88 │        64.31 / 64.76 ±0.35 / 65.18 ms │        62.78 / 64.80 ±3.11 / 70.97 ms │     no change │
│ QQuery 89 │        36.32 / 37.12 ±0.60 / 38.15 ms │        36.16 / 37.18 ±0.62 / 38.10 ms │     no change │
│ QQuery 90 │        17.86 / 20.47 ±5.01 / 30.50 ms │        17.22 / 17.45 ±0.15 / 17.66 ms │ +1.17x faster │
│ QQuery 91 │        46.32 / 47.04 ±0.42 / 47.59 ms │        45.16 / 46.72 ±1.87 / 50.29 ms │     no change │
│ QQuery 92 │        30.40 / 31.25 ±0.82 / 32.80 ms │        30.17 / 30.68 ±0.28 / 30.97 ms │     no change │
│ QQuery 93 │        52.55 / 53.27 ±0.73 / 54.50 ms │        51.66 / 52.04 ±0.25 / 52.35 ms │     no change │
│ QQuery 94 │        39.70 / 41.33 ±2.27 / 45.80 ms │        39.62 / 40.95 ±1.19 / 43.12 ms │     no change │
│ QQuery 95 │        87.49 / 88.68 ±1.35 / 91.31 ms │        85.11 / 86.72 ±1.17 / 88.62 ms │     no change │
│ QQuery 96 │        24.74 / 24.92 ±0.11 / 25.06 ms │        24.42 / 24.68 ±0.17 / 24.91 ms │     no change │
│ QQuery 97 │        46.67 / 47.43 ±0.52 / 48.10 ms │        46.28 / 47.00 ±0.80 / 48.47 ms │     no change │
│ QQuery 98 │        44.12 / 45.56 ±1.05 / 47.25 ms │        43.38 / 44.59 ±1.47 / 47.41 ms │     no change │
│ QQuery 99 │        71.26 / 73.02 ±2.79 / 78.59 ms │        71.49 / 72.43 ±1.44 / 75.28 ms │     no change │
└───────────┴───────────────────────────────────────┴───────────────────────────────────────┴───────────────┘
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━┓
┃ Benchmark Summary                  ┃            ┃
┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━┩
│ Total Time (HEAD)                  │ 10736.94ms │
│ Total Time (alamb_test_coalesce)   │ 10489.03ms │
│ Average Time (HEAD)                │   108.45ms │
│ Average Time (alamb_test_coalesce) │   105.95ms │
│ Queries Faster                     │          9 │
│ Queries Slower                     │          1 │
│ Queries with No Change             │         89 │
│ Queries with Failure               │          0 │
└────────────────────────────────────┴────────────┘

Resource Usage

tpcds — base (merge-base)

Metric	Value
Wall time	55.0s
Peak memory	1.8 GiB
Avg memory	1.3 GiB
CPU user	241.7s
CPU sys	6.5s
Peak spill	0 B

tpcds — branch

Metric	Value
Wall time	55.0s
Peak memory	2.2 GiB
Avg memory	1.5 GiB
CPU user	237.9s
CPU sys	5.9s
Peak spill	0 B

---

File an issue against this benchmark runner

[adriangbot]

🤖 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

Comparing HEAD and alamb_test_coalesce
--------------------
Benchmark clickbench_partitioned.json
--------------------
┏━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓
┃ Query     ┃                                  HEAD ┃                   alamb_test_coalesce ┃        Change ┃
┡━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩
│ QQuery 0  │          1.17 / 3.27 ±4.11 / 11.49 ms │          1.15 / 3.25 ±4.14 / 11.54 ms │     no change │
│ QQuery 1  │        12.13 / 12.43 ±0.18 / 12.63 ms │        12.04 / 12.39 ±0.27 / 12.75 ms │     no change │
│ QQuery 2  │        35.32 / 35.56 ±0.21 / 35.85 ms │        35.34 / 35.72 ±0.27 / 36.11 ms │     no change │
│ QQuery 3  │        30.18 / 30.60 ±0.63 / 31.84 ms │        29.97 / 30.32 ±0.48 / 31.26 ms │     no change │
│ QQuery 4  │     220.32 / 224.71 ±2.87 / 228.69 ms │     220.10 / 224.67 ±2.78 / 228.05 ms │     no change │
│ QQuery 5  │     269.92 / 271.87 ±2.31 / 275.84 ms │     269.62 / 272.62 ±3.78 / 280.07 ms │     no change │
│ QQuery 6  │           1.19 / 1.34 ±0.23 / 1.80 ms │           1.21 / 1.34 ±0.21 / 1.76 ms │     no change │
│ QQuery 7  │        13.18 / 13.44 ±0.21 / 13.71 ms │        13.40 / 13.53 ±0.10 / 13.66 ms │     no change │
│ QQuery 8  │     318.30 / 321.61 ±4.62 / 330.76 ms │     319.81 / 322.59 ±3.25 / 327.89 ms │     no change │
│ QQuery 9  │     450.30 / 458.09 ±7.03 / 468.18 ms │    439.55 / 457.99 ±10.32 / 468.20 ms │     no change │
│ QQuery 10 │        68.42 / 69.15 ±0.72 / 70.32 ms │        67.35 / 68.72 ±0.83 / 69.59 ms │     no change │
│ QQuery 11 │        79.06 / 79.69 ±0.74 / 80.87 ms │       79.00 / 85.11 ±8.36 / 101.61 ms │  1.07x slower │
│ QQuery 12 │     266.24 / 270.21 ±3.36 / 274.15 ms │    266.97 / 278.25 ±10.98 / 297.51 ms │     no change │
│ QQuery 13 │    364.11 / 373.95 ±10.12 / 393.17 ms │    361.21 / 374.73 ±13.07 / 399.45 ms │     no change │
│ QQuery 14 │     276.02 / 282.59 ±6.05 / 289.87 ms │     277.94 / 285.37 ±6.47 / 295.51 ms │     no change │
│ QQuery 15 │     266.10 / 271.68 ±5.03 / 280.29 ms │     267.21 / 277.46 ±8.34 / 288.01 ms │     no change │
│ QQuery 16 │     608.77 / 613.16 ±3.36 / 617.37 ms │     609.98 / 619.30 ±6.49 / 627.21 ms │     no change │
│ QQuery 17 │     613.50 / 619.50 ±6.32 / 630.26 ms │     613.85 / 622.99 ±9.60 / 641.46 ms │     no change │
│ QQuery 18 │ 1251.92 / 1266.21 ±10.43 / 1283.40 ms │ 1254.30 / 1277.88 ±20.43 / 1315.25 ms │     no change │
│ QQuery 19 │        27.01 / 33.07 ±6.96 / 44.01 ms │        27.26 / 27.81 ±0.48 / 28.37 ms │ +1.19x faster │
│ QQuery 20 │     512.19 / 517.48 ±6.86 / 531.06 ms │     510.44 / 517.30 ±6.62 / 527.45 ms │     no change │
│ QQuery 21 │     510.01 / 515.77 ±5.72 / 526.01 ms │     512.10 / 515.30 ±2.55 / 519.23 ms │     no change │
│ QQuery 22 │   980.18 / 992.46 ±10.44 / 1011.42 ms │     981.68 / 989.14 ±6.22 / 998.51 ms │     no change │
│ QQuery 23 │ 3124.06 / 3187.54 ±35.53 / 3232.67 ms │ 3162.42 / 3183.10 ±19.39 / 3215.08 ms │     no change │
│ QQuery 24 │        41.25 / 42.73 ±1.95 / 46.57 ms │       41.19 / 48.68 ±13.84 / 76.32 ms │  1.14x slower │
│ QQuery 25 │     109.29 / 110.07 ±0.91 / 111.83 ms │     109.68 / 113.16 ±4.24 / 121.28 ms │     no change │
│ QQuery 26 │        41.64 / 41.96 ±0.38 / 42.71 ms │        41.27 / 42.22 ±0.94 / 43.61 ms │     no change │
│ QQuery 27 │     671.97 / 677.49 ±7.59 / 692.49 ms │     667.81 / 671.09 ±4.34 / 679.19 ms │     no change │
│ QQuery 28 │ 3002.07 / 3039.69 ±20.34 / 3057.91 ms │ 3000.96 / 3022.32 ±15.23 / 3043.04 ms │     no change │
│ QQuery 29 │        40.63 / 44.18 ±6.10 / 56.34 ms │        39.71 / 45.98 ±7.61 / 57.57 ms │     no change │
│ QQuery 30 │    298.73 / 317.48 ±26.49 / 369.82 ms │     294.65 / 304.22 ±5.73 / 312.04 ms │     no change │
│ QQuery 31 │     276.68 / 286.03 ±7.32 / 294.40 ms │     275.11 / 288.76 ±7.42 / 297.33 ms │     no change │
│ QQuery 32 │    938.65 / 955.26 ±14.74 / 977.67 ms │    916.58 / 942.89 ±19.34 / 976.50 ms │     no change │
│ QQuery 33 │ 1441.85 / 1456.89 ±16.11 / 1485.72 ms │ 1430.93 / 1462.88 ±41.85 / 1545.76 ms │     no change │
│ QQuery 34 │ 1455.12 / 1507.11 ±33.98 / 1562.08 ms │ 1471.60 / 1496.11 ±16.90 / 1522.01 ms │     no change │
│ QQuery 35 │    276.57 / 301.35 ±26.45 / 349.05 ms │    269.59 / 286.28 ±12.32 / 306.74 ms │ +1.05x faster │
│ QQuery 36 │        67.92 / 72.84 ±6.67 / 85.83 ms │        64.60 / 68.87 ±2.75 / 71.93 ms │ +1.06x faster │
│ QQuery 37 │        36.64 / 40.52 ±3.02 / 45.71 ms │        34.99 / 36.87 ±2.88 / 42.46 ms │ +1.10x faster │
│ QQuery 38 │        39.67 / 44.21 ±5.89 / 55.65 ms │        43.38 / 48.50 ±4.36 / 55.53 ms │  1.10x slower │
│ QQuery 39 │     137.65 / 151.42 ±7.55 / 160.09 ms │     133.16 / 141.55 ±4.42 / 145.67 ms │ +1.07x faster │
│ QQuery 40 │        13.37 / 15.84 ±3.98 / 23.77 ms │        13.18 / 14.85 ±2.18 / 19.15 ms │ +1.07x faster │
│ QQuery 41 │        13.16 / 14.57 ±2.43 / 19.42 ms │        13.30 / 14.90 ±2.12 / 19.03 ms │     no change │
│ QQuery 42 │        12.63 / 12.85 ±0.25 / 13.34 ms │        12.89 / 15.98 ±4.15 / 23.76 ms │  1.24x slower │
└───────────┴───────────────────────────────────────┴───────────────────────────────────────┴───────────────┘
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━┓
┃ Benchmark Summary                  ┃            ┃
┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━┩
│ Total Time (HEAD)                  │ 19597.86ms │
│ Total Time (alamb_test_coalesce)   │ 19562.98ms │
│ Average Time (HEAD)                │   455.76ms │
│ Average Time (alamb_test_coalesce) │   454.95ms │
│ Queries Faster                     │          6 │
│ Queries Slower                     │          4 │
│ Queries with No Change             │         33 │
│ Queries with Failure               │          0 │
└────────────────────────────────────┴────────────┘

Resource Usage

clickbench_partitioned — base (merge-base)

Metric	Value
Wall time	100.0s
Peak memory	11.6 GiB
Avg memory	4.5 GiB
CPU user	1003.1s
CPU sys	69.2s
Peak spill	0 B

clickbench_partitioned — branch

Metric	Value
Wall time	100.0s
Peak memory	11.4 GiB
Avg memory	4.7 GiB
CPU user	1002.2s
CPU sys	70.0s
Peak spill	0 B

---

File an issue against this benchmark runner

[alamb]

Performacne looks as good or better in my opinion

[alamb]

run benchmark tpch_sf1

[Dandandan]

run benchmark tpch10

[adriangbot]

🤖 Benchmark running (GKE) | trigger
Instance: c4a-highmem-16 (12 vCPU / 65 GiB) | Linux bench-c4663131111-519-wjtkf 6.12.68+ #1 SMP Sat May 2 07:49:07 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 alamb/test_coalesce (5e19ac2) to 8995ce6 (merge-base) diff using: tpch_sf1
Results will be posted here when complete

---

File an issue against this benchmark runner

[adriangbot]

🤖 Benchmark running (GKE) | trigger
Instance: c4a-highmem-16 (12 vCPU / 65 GiB) | Linux bench-c4663143220-520-m9947 6.12.68+ #1 SMP Sat May 2 07:49:07 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 alamb/test_coalesce (5e19ac2) to 8995ce6 (merge-base) diff using: tpch10
Results will be posted here when complete

---

File an issue against this benchmark runner

[adriangbot]

Benchmark for this request failed.

Last 20 lines of output:

Click to expand

# Micro-Benchmarks (specific operators and features)
cancellation:           How long cancelling a query takes
nlj:                    Benchmark for simple nested loop joins, testing various join scenarios
hj:                     Benchmark for simple hash joins, testing various join scenarios
smj:                    Benchmark for simple sort merge joins, testing various join scenarios
dict:                   Benchmark for dictionary-encoded group-by scenarios
compile_profile:        Compile and execute TPC-H across selected Cargo profiles, reporting timing and binary size


━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
Supported Configuration (Environment Variables)
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
DATA_DIR            directory to store datasets
CARGO_COMMAND       command that runs the benchmark binary
DATAFUSION_DIR      directory to use (default /workspace/datafusion-base)
RESULTS_NAME        folder where the benchmark files are stored
PREFER_HASH_JOIN    Prefer hash join algorithm (default true)
SIMULATE_LATENCY    Simulate object store latency to mimic S3 (default false)
DATAFUSION_*        Set the given datafusion configuration

---

File an issue against this benchmark runner

[adriangbot]

🤖 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

Comparing HEAD and alamb_test_coalesce
--------------------
Benchmark tpch_sf10.json
--------------------
┏━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓
┃ Query     ┃                               HEAD ┃                alamb_test_coalesce ┃        Change ┃
┡━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩
│ QQuery 1  │  307.35 / 310.80 ±2.00 / 312.61 ms │  308.79 / 311.59 ±1.76 / 314.32 ms │     no change │
│ QQuery 2  │  102.24 / 104.66 ±2.47 / 108.95 ms │  102.38 / 106.34 ±2.41 / 109.32 ms │     no change │
│ QQuery 3  │  235.41 / 237.16 ±2.04 / 241.12 ms │  232.59 / 236.20 ±2.42 / 239.22 ms │     no change │
│ QQuery 4  │  116.44 / 118.35 ±2.13 / 121.80 ms │  112.17 / 113.56 ±1.00 / 115.06 ms │     no change │
│ QQuery 5  │  349.39 / 358.67 ±5.89 / 367.31 ms │  353.80 / 358.54 ±4.04 / 364.28 ms │     no change │
│ QQuery 6  │  124.29 / 128.50 ±3.74 / 134.06 ms │  124.64 / 126.36 ±2.45 / 131.18 ms │     no change │
│ QQuery 7  │ 453.29 / 466.41 ±10.52 / 484.09 ms │  458.08 / 468.19 ±8.16 / 478.27 ms │     no change │
│ QQuery 8  │  377.12 / 385.47 ±6.86 / 397.00 ms │  382.79 / 391.42 ±8.40 / 405.93 ms │     no change │
│ QQuery 9  │  546.05 / 554.61 ±4.59 / 559.57 ms │  537.52 / 554.02 ±9.48 / 565.36 ms │     no change │
│ QQuery 10 │ 307.58 / 324.29 ±14.83 / 343.42 ms │ 301.29 / 316.24 ±11.76 / 332.93 ms │     no change │
│ QQuery 11 │   83.85 / 95.40 ±11.46 / 112.79 ms │    86.15 / 93.83 ±9.63 / 109.75 ms │     no change │
│ QQuery 12 │  177.47 / 183.12 ±4.96 / 192.14 ms │  178.15 / 181.99 ±3.89 / 189.01 ms │     no change │
│ QQuery 13 │  282.34 / 294.53 ±9.27 / 308.09 ms │  283.81 / 292.82 ±5.43 / 300.48 ms │     no change │
│ QQuery 14 │  173.49 / 176.10 ±2.17 / 180.06 ms │  172.68 / 178.20 ±5.57 / 188.46 ms │     no change │
│ QQuery 15 │  304.07 / 309.73 ±4.81 / 317.95 ms │  305.20 / 307.10 ±1.49 / 308.92 ms │     no change │
│ QQuery 16 │     65.80 / 67.64 ±1.94 / 71.32 ms │     64.75 / 67.30 ±1.57 / 69.17 ms │     no change │
│ QQuery 17 │  636.21 / 641.73 ±4.64 / 648.40 ms │ 624.86 / 633.93 ±10.51 / 652.22 ms │     no change │
│ QQuery 18 │ 687.49 / 696.77 ±11.81 / 718.77 ms │ 665.08 / 685.17 ±13.96 / 708.88 ms │     no change │
│ QQuery 19 │ 246.86 / 262.21 ±16.30 / 283.60 ms │ 243.20 / 260.68 ±14.42 / 276.27 ms │     no change │
│ QQuery 20 │ 276.95 / 292.16 ±11.12 / 310.48 ms │  280.28 / 284.37 ±4.95 / 293.46 ms │     no change │
│ QQuery 21 │ 777.62 / 795.96 ±15.95 / 819.76 ms │  784.96 / 795.27 ±8.42 / 807.12 ms │     no change │
│ QQuery 22 │   61.20 / 87.31 ±34.80 / 153.23 ms │    60.45 / 75.08 ±11.87 / 92.07 ms │ +1.16x faster │
└───────────┴────────────────────────────────────┴────────────────────────────────────┴───────────────┘
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━┓
┃ Benchmark Summary                  ┃           ┃
┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━┩
│ Total Time (HEAD)                  │ 6891.58ms │
│ Total Time (alamb_test_coalesce)   │ 6838.20ms │
│ Average Time (HEAD)                │  313.25ms │
│ Average Time (alamb_test_coalesce) │  310.83ms │
│ Queries Faster                     │         1 │
│ Queries Slower                     │         0 │
│ Queries with No Change             │        21 │
│ Queries with Failure               │         0 │
└────────────────────────────────────┴───────────┘

Resource Usage

tpch10 — base (merge-base)

Metric	Value
Wall time	35.0s
Peak memory	4.5 GiB
Avg memory	1.6 GiB
CPU user	343.1s
CPU sys	20.4s
Peak spill	0 B

tpch10 — branch

Metric	Value
Wall time	35.0s
Peak memory	4.5 GiB
Avg memory	1.7 GiB
CPU user	343.3s
CPU sys	20.3s
Peak spill	0 B

---

File an issue against this benchmark runner