feat: Optimise convert_to_state for SUM and BIT_OR_XOR

[Mark1626]

Which issue does this PR close?

Rationale for this change

I have a query where GroupedHashAggregateStream was switching to SkippingAggregation. I noticed convert_to_state coming up as a bottleneck, particular the memory allocation and the arrow kernel

What changes are included in this PR?

Add a fast path for sum and bit OR/XOR, where convert_to_state returns after applying the filter

Are these changes tested?

Yes

Are there any user-facing changes?

No

[Dandandan]

run benchmarks

[adriangbot]

🤖 Benchmark running (GKE) | trigger
Instance: c4a-highmem-16 (12 vCPU / 65 GiB) | Linux bench-c4214049145-1015-zdk7t 6.12.55+ #1 SMP Sun Feb 1 08:59:41 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 perf-convert-to-state (23fc979) to e1ad871 (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-c4214049145-1016-pqjmq 6.12.55+ #1 SMP Sun Feb 1 08:59:41 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 perf-convert-to-state (23fc979) to e1ad871 (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-c4214049145-1014-jpdmc 6.12.55+ #1 SMP Sun Feb 1 08:59:41 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 perf-convert-to-state (23fc979) to e1ad871 (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 perf-convert-to-state
--------------------
Benchmark tpcds_sf1.json
--------------------
┏━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓
┃ Query     ┃                                     HEAD ┃                    perf-convert-to-state ┃        Change ┃
┡━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩
│ QQuery 1  │              6.53 / 7.00 ±0.78 / 8.55 ms │              6.48 / 6.94 ±0.78 / 8.50 ms │     no change │
│ QQuery 2  │        145.18 / 145.44 ±0.26 / 145.81 ms │        143.51 / 144.46 ±0.95 / 146.06 ms │     no change │
│ QQuery 3  │        111.88 / 113.35 ±1.15 / 115.40 ms │        111.92 / 112.72 ±0.66 / 113.67 ms │     no change │
│ QQuery 4  │    1251.18 / 1271.32 ±12.24 / 1289.41 ms │    1255.64 / 1275.23 ±15.65 / 1296.50 ms │     no change │
│ QQuery 5  │        171.01 / 172.79 ±2.14 / 176.66 ms │        170.79 / 171.81 ±0.99 / 173.54 ms │     no change │
│ QQuery 6  │       804.91 / 823.68 ±17.05 / 850.86 ms │       821.87 / 847.08 ±21.23 / 877.60 ms │     no change │
│ QQuery 7  │        337.39 / 339.78 ±1.33 / 341.35 ms │        337.93 / 339.31 ±1.39 / 341.10 ms │     no change │
│ QQuery 8  │        115.05 / 116.01 ±0.80 / 117.06 ms │        113.46 / 115.80 ±1.92 / 118.51 ms │     no change │
│ QQuery 9  │        100.38 / 102.83 ±1.94 / 106.14 ms │        101.24 / 107.46 ±5.33 / 117.38 ms │     no change │
│ QQuery 10 │        105.50 / 106.79 ±0.92 / 107.90 ms │        105.29 / 105.69 ±0.27 / 106.03 ms │     no change │
│ QQuery 11 │       862.46 / 876.64 ±10.52 / 892.86 ms │        859.65 / 868.21 ±6.38 / 876.60 ms │     no change │
│ QQuery 12 │           44.31 / 46.77 ±1.79 / 49.61 ms │           43.94 / 45.31 ±1.06 / 47.10 ms │     no change │
│ QQuery 13 │        395.36 / 399.57 ±3.59 / 404.65 ms │        397.20 / 398.71 ±1.22 / 400.57 ms │     no change │
│ QQuery 14 │       988.10 / 992.49 ±5.50 / 1003.05 ms │       983.01 / 990.63 ±6.45 / 1001.38 ms │     no change │
│ QQuery 15 │           14.75 / 16.38 ±1.23 / 18.22 ms │           15.55 / 16.46 ±1.05 / 18.43 ms │     no change │
│ QQuery 16 │              6.93 / 7.84 ±0.94 / 9.08 ms │             7.42 / 8.17 ±1.21 / 10.59 ms │     no change │
│ QQuery 17 │        224.68 / 226.07 ±0.86 / 227.08 ms │        224.71 / 226.52 ±1.11 / 228.07 ms │     no change │
│ QQuery 18 │        126.08 / 127.34 ±1.31 / 129.64 ms │        124.72 / 125.68 ±0.76 / 127.02 ms │     no change │
│ QQuery 19 │        153.96 / 155.33 ±0.77 / 156.36 ms │        151.93 / 153.40 ±1.38 / 155.66 ms │     no change │
│ QQuery 20 │           13.23 / 13.76 ±0.38 / 14.24 ms │           13.06 / 13.75 ±0.71 / 15.11 ms │     no change │
│ QQuery 21 │           18.54 / 19.00 ±0.34 / 19.45 ms │           18.43 / 18.98 ±0.38 / 19.52 ms │     no change │
│ QQuery 22 │        478.19 / 482.57 ±2.78 / 486.21 ms │        479.13 / 483.99 ±3.29 / 488.24 ms │     no change │
│ QQuery 23 │        845.44 / 854.70 ±6.16 / 860.83 ms │        853.51 / 861.37 ±7.06 / 871.53 ms │     no change │
│ QQuery 24 │        375.99 / 378.01 ±1.66 / 380.92 ms │        372.87 / 377.76 ±4.55 / 386.40 ms │     no change │
│ QQuery 25 │        335.73 / 338.45 ±1.76 / 340.62 ms │        334.60 / 335.99 ±1.41 / 338.09 ms │     no change │
│ QQuery 26 │           81.19 / 82.29 ±0.76 / 83.22 ms │           79.78 / 81.67 ±1.45 / 83.95 ms │     no change │
│ QQuery 27 │              6.91 / 7.20 ±0.22 / 7.52 ms │              6.82 / 7.72 ±0.96 / 8.95 ms │  1.07x slower │
│ QQuery 28 │        147.91 / 149.07 ±1.29 / 151.56 ms │        148.43 / 149.36 ±1.08 / 151.43 ms │     no change │
│ QQuery 29 │        278.65 / 279.77 ±0.70 / 280.87 ms │        276.36 / 277.70 ±1.16 / 279.58 ms │     no change │
│ QQuery 30 │           43.65 / 45.36 ±1.54 / 48.14 ms │           42.39 / 44.06 ±1.26 / 46.29 ms │     no change │
│ QQuery 31 │        168.10 / 170.02 ±1.20 / 171.43 ms │        167.42 / 169.23 ±1.57 / 172.03 ms │     no change │
│ QQuery 32 │           57.49 / 58.44 ±0.88 / 59.71 ms │           56.38 / 57.19 ±0.73 / 58.43 ms │     no change │
│ QQuery 33 │        138.59 / 139.64 ±1.10 / 141.43 ms │        138.80 / 140.23 ±1.29 / 142.22 ms │     no change │
│ QQuery 34 │              6.83 / 7.29 ±0.46 / 8.15 ms │              6.77 / 7.02 ±0.32 / 7.64 ms │     no change │
│ QQuery 35 │        105.76 / 107.42 ±1.09 / 109.04 ms │        106.43 / 107.80 ±1.15 / 109.19 ms │     no change │
│ QQuery 36 │              6.41 / 6.70 ±0.25 / 7.06 ms │              6.56 / 6.71 ±0.15 / 6.96 ms │     no change │
│ QQuery 37 │              8.00 / 8.47 ±0.35 / 8.94 ms │             7.97 / 9.06 ±1.25 / 11.46 ms │  1.07x slower │
│ QQuery 38 │           83.03 / 85.44 ±1.77 / 88.41 ms │           82.75 / 86.77 ±3.27 / 92.15 ms │     no change │
│ QQuery 39 │        120.91 / 123.94 ±1.93 / 125.74 ms │        120.19 / 124.66 ±3.38 / 128.40 ms │     no change │
│ QQuery 40 │        107.59 / 114.60 ±5.50 / 124.17 ms │        106.07 / 113.59 ±6.14 / 123.45 ms │     no change │
│ QQuery 41 │           14.60 / 16.11 ±1.16 / 18.10 ms │           14.26 / 14.81 ±0.36 / 15.17 ms │ +1.09x faster │
│ QQuery 42 │        107.21 / 107.95 ±0.70 / 108.97 ms │        106.53 / 108.17 ±1.15 / 110.11 ms │     no change │
│ QQuery 43 │              5.76 / 5.94 ±0.18 / 6.29 ms │              6.01 / 6.12 ±0.16 / 6.44 ms │     no change │
│ QQuery 44 │           11.71 / 12.22 ±0.59 / 13.28 ms │           11.64 / 12.71 ±0.99 / 14.32 ms │     no change │
│ QQuery 45 │           49.38 / 50.06 ±0.68 / 50.91 ms │           49.94 / 50.63 ±0.57 / 51.49 ms │     no change │
│ QQuery 46 │              8.17 / 8.40 ±0.16 / 8.65 ms │              8.36 / 8.48 ±0.11 / 8.60 ms │     no change │
│ QQuery 47 │        669.66 / 678.85 ±4.96 / 684.25 ms │        675.20 / 679.76 ±2.36 / 681.74 ms │     no change │
│ QQuery 48 │        281.01 / 288.36 ±5.01 / 295.64 ms │        277.32 / 289.48 ±6.27 / 294.48 ms │     no change │
│ QQuery 49 │        248.29 / 253.36 ±4.11 / 258.94 ms │        249.50 / 251.14 ±1.60 / 253.26 ms │     no change │
│ QQuery 50 │        221.29 / 223.57 ±1.87 / 226.95 ms │        219.92 / 223.80 ±3.19 / 229.55 ms │     no change │
│ QQuery 51 │        176.66 / 181.16 ±2.89 / 185.17 ms │        180.51 / 181.60 ±0.81 / 183.01 ms │     no change │
│ QQuery 52 │        106.09 / 107.24 ±0.76 / 108.45 ms │        105.77 / 106.81 ±0.75 / 108.10 ms │     no change │
│ QQuery 53 │        100.46 / 101.36 ±0.66 / 102.40 ms │        100.43 / 102.11 ±1.87 / 105.60 ms │     no change │
│ QQuery 54 │        143.85 / 145.09 ±0.95 / 146.67 ms │        142.57 / 144.70 ±1.67 / 147.28 ms │     no change │
│ QQuery 55 │        104.80 / 106.60 ±1.11 / 107.89 ms │        105.04 / 105.54 ±0.40 / 106.20 ms │     no change │
│ QQuery 56 │        139.34 / 140.86 ±1.97 / 144.64 ms │        137.59 / 139.90 ±2.42 / 144.42 ms │     no change │
│ QQuery 57 │        171.79 / 173.43 ±1.66 / 176.31 ms │        170.44 / 172.28 ±1.25 / 173.71 ms │     no change │
│ QQuery 58 │        284.99 / 292.64 ±6.62 / 304.33 ms │        285.52 / 291.30 ±4.36 / 297.58 ms │     no change │
│ QQuery 59 │        197.34 / 203.64 ±8.60 / 220.55 ms │        194.79 / 196.01 ±0.92 / 196.98 ms │     no change │
│ QQuery 60 │        142.50 / 143.56 ±0.57 / 144.14 ms │        140.17 / 142.39 ±1.55 / 144.33 ms │     no change │
│ QQuery 61 │           12.91 / 13.10 ±0.21 / 13.51 ms │           12.83 / 13.02 ±0.24 / 13.49 ms │     no change │
│ QQuery 62 │       911.76 / 930.57 ±18.76 / 964.68 ms │       887.02 / 911.99 ±18.98 / 940.38 ms │     no change │
│ QQuery 63 │        102.78 / 105.78 ±2.62 / 109.82 ms │        101.13 / 105.00 ±2.57 / 108.96 ms │     no change │
│ QQuery 64 │        671.70 / 677.66 ±3.07 / 680.33 ms │        673.81 / 675.84 ±2.00 / 678.68 ms │     no change │
│ QQuery 65 │        246.57 / 248.07 ±1.40 / 250.44 ms │        247.34 / 251.09 ±2.65 / 255.13 ms │     no change │
│ QQuery 66 │       228.92 / 253.51 ±16.58 / 278.04 ms │       238.84 / 251.61 ±14.11 / 274.77 ms │     no change │
│ QQuery 67 │        301.79 / 313.03 ±5.98 / 319.46 ms │        306.18 / 311.12 ±3.92 / 315.60 ms │     no change │
│ QQuery 68 │            8.59 / 10.94 ±1.48 / 12.75 ms │             8.42 / 9.65 ±0.67 / 10.24 ms │ +1.13x faster │
│ QQuery 69 │        101.23 / 102.98 ±1.60 / 105.76 ms │        100.98 / 102.44 ±1.11 / 104.20 ms │     no change │
│ QQuery 70 │        341.10 / 351.26 ±9.67 / 369.59 ms │       315.89 / 336.65 ±12.33 / 354.27 ms │     no change │
│ QQuery 71 │        134.99 / 139.69 ±6.82 / 153.24 ms │        132.73 / 134.45 ±1.48 / 137.00 ms │     no change │
│ QQuery 72 │        609.93 / 617.88 ±7.87 / 628.99 ms │        600.22 / 612.83 ±6.87 / 619.91 ms │     no change │
│ QQuery 73 │              7.45 / 8.16 ±0.50 / 8.96 ms │              6.49 / 7.32 ±0.72 / 8.54 ms │ +1.12x faster │
│ QQuery 74 │       543.27 / 578.46 ±22.61 / 605.27 ms │        533.65 / 540.08 ±5.78 / 549.01 ms │ +1.07x faster │
│ QQuery 75 │        274.60 / 276.66 ±1.81 / 279.14 ms │        271.85 / 273.97 ±2.03 / 277.67 ms │     no change │
│ QQuery 76 │        131.71 / 133.56 ±1.38 / 134.75 ms │        129.67 / 131.31 ±1.74 / 134.46 ms │     no change │
│ QQuery 77 │        187.01 / 188.42 ±1.35 / 190.65 ms │        183.93 / 186.17 ±1.54 / 188.03 ms │     no change │
│ QQuery 78 │        333.31 / 338.03 ±2.64 / 341.06 ms │        333.18 / 337.94 ±2.49 / 340.39 ms │     no change │
│ QQuery 79 │        227.77 / 229.71 ±2.70 / 234.98 ms │        228.21 / 231.24 ±2.24 / 234.06 ms │     no change │
│ QQuery 80 │        322.32 / 324.20 ±1.31 / 325.98 ms │        318.30 / 320.91 ±3.00 / 326.43 ms │     no change │
│ QQuery 81 │           25.67 / 27.27 ±1.39 / 29.76 ms │           25.55 / 26.21 ±0.55 / 27.22 ms │     no change │
│ QQuery 82 │        197.72 / 200.58 ±2.23 / 202.88 ms │        197.22 / 200.13 ±2.28 / 202.83 ms │     no change │
│ QQuery 83 │           38.07 / 38.88 ±0.80 / 40.21 ms │           37.59 / 38.55 ±1.37 / 41.26 ms │     no change │
│ QQuery 84 │           47.99 / 48.73 ±1.02 / 50.75 ms │           47.75 / 48.53 ±0.67 / 49.51 ms │     no change │
│ QQuery 85 │        147.17 / 147.79 ±0.95 / 149.68 ms │        145.29 / 146.94 ±1.17 / 148.44 ms │     no change │
│ QQuery 86 │           39.22 / 39.99 ±0.84 / 41.59 ms │           37.29 / 38.70 ±1.19 / 40.61 ms │     no change │
│ QQuery 87 │           83.97 / 88.18 ±3.01 / 93.14 ms │           83.27 / 86.25 ±3.35 / 92.71 ms │     no change │
│ QQuery 88 │          98.85 / 99.69 ±1.04 / 101.72 ms │           98.13 / 99.18 ±0.56 / 99.75 ms │     no change │
│ QQuery 89 │        117.08 / 118.81 ±1.31 / 120.50 ms │        117.78 / 118.68 ±0.88 / 119.97 ms │     no change │
│ QQuery 90 │           23.36 / 23.94 ±0.40 / 24.54 ms │           22.95 / 23.36 ±0.35 / 23.87 ms │     no change │
│ QQuery 91 │           61.82 / 63.27 ±1.01 / 64.48 ms │           61.81 / 63.49 ±0.85 / 64.01 ms │     no change │
│ QQuery 92 │           56.42 / 57.43 ±0.94 / 59.21 ms │           55.20 / 56.85 ±1.08 / 58.24 ms │     no change │
│ QQuery 93 │        182.84 / 186.95 ±2.69 / 190.86 ms │        181.47 / 184.16 ±1.94 / 186.43 ms │     no change │
│ QQuery 94 │           60.79 / 61.42 ±0.43 / 62.03 ms │           60.36 / 61.37 ±0.83 / 62.54 ms │     no change │
│ QQuery 95 │        125.28 / 127.79 ±1.88 / 130.54 ms │        124.55 / 125.57 ±0.64 / 126.54 ms │     no change │
│ QQuery 96 │           72.29 / 73.49 ±0.96 / 74.69 ms │           72.89 / 73.72 ±0.72 / 75.01 ms │     no change │
│ QQuery 97 │        123.24 / 126.66 ±1.85 / 128.23 ms │        122.61 / 126.67 ±2.10 / 128.55 ms │     no change │
│ QQuery 98 │        150.86 / 153.01 ±1.95 / 156.12 ms │        151.78 / 152.90 ±1.24 / 155.01 ms │     no change │
│ QQuery 99 │ 10710.29 / 10760.14 ±36.73 / 10817.60 ms │ 10695.94 / 10718.27 ±22.50 / 10751.34 ms │     no change │
└───────────┴──────────────────────────────────────────┴──────────────────────────────────────────┴───────────────┘
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━┓
┃ Benchmark Summary                    ┃            ┃
┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━┩
│ Total Time (HEAD)                    │ 31045.61ms │
│ Total Time (perf-convert-to-state)   │ 30896.08ms │
│ Average Time (HEAD)                  │   313.59ms │
│ Average Time (perf-convert-to-state) │   312.08ms │
│ Queries Faster                       │          4 │
│ Queries Slower                       │          2 │
│ Queries with No Change               │         93 │
│ Queries with Failure                 │          0 │
└──────────────────────────────────────┴────────────┘

Resource Usage

tpcds — base (merge-base)

Metric	Value
Wall time	155.5s
Peak memory	5.5 GiB
Avg memory	4.4 GiB
CPU user	256.1s
CPU sys	16.4s
Peak spill	0 B

tpcds — branch

Metric	Value
Wall time	154.8s
Peak memory	5.7 GiB
Avg memory	4.7 GiB
CPU user	255.2s
CPU sys	15.9s
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 perf-convert-to-state
--------------------
Benchmark clickbench_partitioned.json
--------------------
┏━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓
┃ Query     ┃                                  HEAD ┃                 perf-convert-to-state ┃        Change ┃
┡━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩
│ QQuery 0  │          1.20 / 4.47 ±6.39 / 17.26 ms │          1.21 / 4.41 ±6.33 / 17.06 ms │     no change │
│ QQuery 1  │        14.18 / 14.57 ±0.21 / 14.77 ms │        14.27 / 14.61 ±0.19 / 14.81 ms │     no change │
│ QQuery 2  │        43.51 / 43.80 ±0.19 / 43.98 ms │        43.06 / 43.68 ±0.38 / 44.07 ms │     no change │
│ QQuery 3  │        41.29 / 43.31 ±2.08 / 46.97 ms │        41.43 / 44.39 ±1.71 / 46.78 ms │     no change │
│ QQuery 4  │     287.28 / 293.85 ±6.28 / 304.94 ms │     284.92 / 297.12 ±8.38 / 309.78 ms │     no change │
│ QQuery 5  │     346.38 / 351.25 ±2.61 / 353.46 ms │     341.38 / 343.47 ±1.15 / 344.81 ms │     no change │
│ QQuery 6  │           5.33 / 6.03 ±0.58 / 7.03 ms │           5.59 / 6.43 ±0.65 / 7.58 ms │  1.07x slower │
│ QQuery 7  │        16.67 / 17.37 ±0.56 / 18.37 ms │        16.82 / 17.83 ±1.13 / 19.91 ms │     no change │
│ QQuery 8  │    427.56 / 445.33 ±13.53 / 468.51 ms │     419.24 / 430.60 ±7.60 / 440.77 ms │     no change │
│ QQuery 9  │     651.84 / 665.42 ±8.78 / 676.06 ms │     649.49 / 656.96 ±5.48 / 662.41 ms │     no change │
│ QQuery 10 │        91.05 / 94.16 ±2.08 / 96.54 ms │       92.60 / 96.73 ±3.93 / 103.36 ms │     no change │
│ QQuery 11 │     103.80 / 104.12 ±0.24 / 104.39 ms │     104.56 / 105.92 ±1.15 / 107.34 ms │     no change │
│ QQuery 12 │     338.20 / 342.26 ±2.90 / 346.16 ms │     343.80 / 353.97 ±8.68 / 365.29 ms │     no change │
│ QQuery 13 │     462.16 / 470.21 ±7.38 / 481.43 ms │    456.56 / 480.37 ±14.49 / 497.39 ms │     no change │
│ QQuery 14 │     345.81 / 350.34 ±3.45 / 354.06 ms │     347.46 / 351.68 ±2.32 / 353.86 ms │     no change │
│ QQuery 15 │     356.68 / 363.82 ±5.60 / 373.01 ms │    358.70 / 387.23 ±38.98 / 464.59 ms │  1.06x slower │
│ QQuery 16 │    707.13 / 733.49 ±23.95 / 769.55 ms │     720.05 / 728.03 ±5.90 / 735.88 ms │     no change │
│ QQuery 17 │     704.31 / 713.12 ±6.62 / 720.98 ms │     707.88 / 716.84 ±7.35 / 726.62 ms │     no change │
│ QQuery 18 │ 1433.11 / 1468.78 ±28.98 / 1501.22 ms │ 1437.72 / 1480.28 ±29.46 / 1527.49 ms │     no change │
│ QQuery 19 │        35.98 / 37.16 ±1.51 / 39.99 ms │      35.88 / 60.62 ±23.95 / 100.42 ms │  1.63x slower │
│ QQuery 20 │    710.76 / 735.67 ±25.74 / 777.77 ms │    712.94 / 737.84 ±32.09 / 800.66 ms │     no change │
│ QQuery 21 │     755.97 / 762.46 ±5.68 / 773.09 ms │     757.72 / 763.94 ±7.12 / 777.55 ms │     no change │
│ QQuery 22 │  1124.91 / 1133.32 ±6.86 / 1143.35 ms │  1135.16 / 1140.58 ±3.91 / 1144.65 ms │     no change │
│ QQuery 23 │ 3062.64 / 3087.04 ±13.81 / 3105.43 ms │ 3062.31 / 3104.21 ±27.50 / 3145.59 ms │     no change │
│ QQuery 24 │      98.65 / 101.45 ±2.02 / 103.94 ms │      99.44 / 101.66 ±1.74 / 104.39 ms │     no change │
│ QQuery 25 │     138.03 / 139.15 ±0.67 / 140.02 ms │     138.84 / 141.78 ±2.65 / 146.20 ms │     no change │
│ QQuery 26 │      96.45 / 100.94 ±3.14 / 105.31 ms │      96.13 / 100.10 ±2.20 / 102.77 ms │     no change │
│ QQuery 27 │     841.46 / 850.98 ±7.50 / 863.90 ms │     843.46 / 849.87 ±8.63 / 866.89 ms │     no change │
│ QQuery 28 │ 7697.79 / 7750.95 ±34.61 / 7797.29 ms │ 7701.25 / 7754.49 ±29.47 / 7785.56 ms │     no change │
│ QQuery 29 │        49.24 / 56.72 ±8.94 / 73.42 ms │     53.84 / 121.61 ±82.23 / 257.06 ms │  2.14x slower │
│ QQuery 30 │     354.37 / 361.13 ±4.36 / 367.21 ms │     365.59 / 379.59 ±7.78 / 387.52 ms │  1.05x slower │
│ QQuery 31 │     357.13 / 372.14 ±9.60 / 381.59 ms │    381.21 / 407.34 ±16.33 / 422.56 ms │  1.09x slower │
│ QQuery 32 │ 1187.18 / 1240.24 ±62.99 / 1355.19 ms │ 1210.94 / 1253.06 ±26.83 / 1295.08 ms │     no change │
│ QQuery 33 │ 1524.83 / 1559.02 ±23.77 / 1585.61 ms │ 1456.62 / 1493.40 ±40.36 / 1568.18 ms │     no change │
│ QQuery 34 │ 1476.01 / 1576.09 ±69.61 / 1656.13 ms │  1456.99 / 1465.78 ±5.90 / 1475.52 ms │ +1.08x faster │
│ QQuery 35 │     385.71 / 391.89 ±6.50 / 401.26 ms │     384.51 / 388.62 ±5.07 / 397.09 ms │     no change │
│ QQuery 36 │     113.04 / 117.31 ±3.21 / 121.17 ms │     120.01 / 124.21 ±5.12 / 134.12 ms │  1.06x slower │
│ QQuery 37 │        48.00 / 50.07 ±1.83 / 52.26 ms │        48.41 / 50.42 ±1.44 / 52.28 ms │     no change │
│ QQuery 38 │        75.46 / 77.51 ±1.42 / 79.69 ms │        76.20 / 77.44 ±0.66 / 77.94 ms │     no change │
│ QQuery 39 │     215.16 / 221.48 ±4.83 / 229.01 ms │     209.12 / 223.43 ±8.53 / 232.31 ms │     no change │
│ QQuery 40 │        24.80 / 26.32 ±0.99 / 27.72 ms │        23.14 / 28.38 ±3.33 / 31.48 ms │  1.08x slower │
│ QQuery 41 │        19.72 / 20.23 ±0.32 / 20.62 ms │        19.24 / 20.18 ±0.54 / 20.72 ms │     no change │
│ QQuery 42 │        19.37 / 20.10 ±0.38 / 20.49 ms │        19.09 / 19.80 ±0.49 / 20.35 ms │     no change │
└───────────┴───────────────────────────────────────┴───────────────────────────────────────┴───────────────┘
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━┓
┃ Benchmark Summary                    ┃            ┃
┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━┩
│ Total Time (HEAD)                    │ 27315.08ms │
│ Total Time (perf-convert-to-state)   │ 27368.87ms │
│ Average Time (HEAD)                  │   635.23ms │
│ Average Time (perf-convert-to-state) │   636.49ms │
│ Queries Faster                       │          1 │
│ Queries Slower                       │          8 │
│ Queries with No Change               │         34 │
│ Queries with Failure                 │          0 │
└──────────────────────────────────────┴────────────┘

Resource Usage

clickbench_partitioned — base (merge-base)

Metric	Value
Wall time	137.6s
Peak memory	35.4 GiB
Avg memory	26.7 GiB
CPU user	1287.0s
CPU sys	103.7s
Peak spill	0 B

clickbench_partitioned — branch

Metric	Value
Wall time	138.0s
Peak memory	41.1 GiB
Avg memory	29.2 GiB
CPU user	1289.5s
CPU sys	104.3s
Peak spill	0 B

---

File an issue against this benchmark runner

[Mark1626]

Not sure why Clickbench Q19 is degrading, it doesn't have an agg

SELECT "UserID" FROM hits WHERE "UserID" = 435090932899640449;

[Dandandan]

run benchmarks

[Dandandan]

Not sure why Clickbench Q19 is degrading, it doesn't have an agg

SELECT "UserID" FROM hits WHERE "UserID" = 435090932899640449;

Some of them are slower sometimes (usually one of the runs). If you look at the minimum time, they are similar.

[Dandandan]

run benchmark tpch tpch10

[adriangbot]

🤖 Benchmark running (GKE) | trigger
Instance: c4a-highmem-16 (12 vCPU / 65 GiB) | Linux bench-c4214997819-1022-lq9sx 6.12.55+ #1 SMP Sun Feb 1 08:59:41 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 perf-convert-to-state (23fc979) to e1ad871 (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-c4214997819-1021-hgdb4 6.12.55+ #1 SMP Sun Feb 1 08:59:41 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 perf-convert-to-state (23fc979) to e1ad871 (merge-base) diff using: clickbench_partitioned
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-c4214997819-1023-mhmck 6.12.55+ #1 SMP Sun Feb 1 08:59:41 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 perf-convert-to-state (23fc979) to e1ad871 (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-c4215004644-1024-g8fr8 6.12.55+ #1 SMP Sun Feb 1 08:59:41 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 perf-convert-to-state (23fc979) to e1ad871 (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-c4215004644-1025-txt2s 6.12.55+ #1 SMP Sun Feb 1 08:59:41 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 perf-convert-to-state (23fc979) to e1ad871 (merge-base) diff using: tpch10
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 perf-convert-to-state
--------------------
Benchmark tpch_sf10.json
--------------------
┏━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓
┃ Query     ┃                               HEAD ┃              perf-convert-to-state ┃        Change ┃
┡━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩
│ QQuery 1  │  370.24 / 376.04 ±5.34 / 385.01 ms │  372.84 / 377.34 ±2.95 / 381.78 ms │     no change │
│ QQuery 2  │  137.73 / 140.64 ±1.78 / 142.96 ms │  137.19 / 143.56 ±4.81 / 151.58 ms │     no change │
│ QQuery 3  │  301.35 / 307.86 ±4.37 / 315.11 ms │  297.72 / 311.34 ±7.25 / 319.00 ms │     no change │
│ QQuery 4  │  160.96 / 161.91 ±0.78 / 163.22 ms │  158.39 / 163.78 ±4.38 / 170.15 ms │     no change │
│ QQuery 5  │  450.88 / 455.69 ±3.75 / 461.84 ms │  446.71 / 458.58 ±7.10 / 466.74 ms │     no change │
│ QQuery 6  │  133.08 / 134.80 ±0.92 / 135.75 ms │  131.58 / 138.61 ±6.05 / 149.13 ms │     no change │
│ QQuery 7  │ 568.12 / 588.96 ±17.85 / 614.22 ms │ 568.76 / 588.33 ±15.85 / 608.78 ms │     no change │
│ QQuery 8  │ 466.75 / 480.86 ±12.46 / 502.74 ms │ 490.87 / 504.06 ±12.28 / 525.27 ms │     no change │
│ QQuery 9  │ 685.58 / 698.11 ±12.68 / 714.49 ms │  712.15 / 717.48 ±8.97 / 735.32 ms │     no change │
│ QQuery 10 │  329.58 / 339.02 ±8.05 / 352.29 ms │  344.55 / 356.26 ±6.01 / 360.42 ms │  1.05x slower │
│ QQuery 11 │  105.67 / 109.27 ±2.36 / 112.49 ms │  109.55 / 112.18 ±1.69 / 114.28 ms │     no change │
│ QQuery 12 │  209.28 / 211.19 ±1.11 / 212.65 ms │  206.59 / 210.82 ±2.81 / 215.26 ms │     no change │
│ QQuery 13 │  347.47 / 353.10 ±5.46 / 361.24 ms │  326.84 / 336.51 ±9.84 / 353.57 ms │     no change │
│ QQuery 14 │  186.22 / 187.93 ±1.29 / 189.39 ms │  188.80 / 190.31 ±1.15 / 191.92 ms │     no change │
│ QQuery 15 │  344.32 / 346.20 ±1.90 / 348.91 ms │  334.87 / 342.42 ±8.14 / 353.20 ms │     no change │
│ QQuery 16 │     91.59 / 92.51 ±0.80 / 93.57 ms │     85.87 / 87.92 ±2.22 / 90.66 ms │     no change │
│ QQuery 17 │ 837.16 / 853.60 ±10.20 / 865.42 ms │ 788.59 / 812.23 ±13.88 / 827.00 ms │     no change │
│ QQuery 18 │ 904.04 / 943.94 ±23.25 / 967.02 ms │ 848.17 / 884.47 ±33.32 / 936.95 ms │ +1.07x faster │
│ QQuery 19 │  279.83 / 282.67 ±3.02 / 287.52 ms │  268.62 / 273.93 ±7.36 / 288.23 ms │     no change │
│ QQuery 20 │  346.25 / 349.55 ±4.41 / 358.13 ms │  321.61 / 328.21 ±5.74 / 338.24 ms │ +1.07x faster │
│ QQuery 21 │ 890.42 / 908.61 ±16.46 / 939.64 ms │ 897.40 / 907.62 ±10.92 / 928.15 ms │     no change │
│ QQuery 22 │     82.13 / 85.83 ±2.50 / 88.70 ms │     83.85 / 86.16 ±1.32 / 87.87 ms │     no change │
└───────────┴────────────────────────────────────┴────────────────────────────────────┴───────────────┘
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━┓
┃ Benchmark Summary                    ┃           ┃
┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━┩
│ Total Time (HEAD)                    │ 8408.31ms │
│ Total Time (perf-convert-to-state)   │ 8332.13ms │
│ Average Time (HEAD)                  │  382.20ms │
│ Average Time (perf-convert-to-state) │  378.73ms │
│ Queries Faster                       │         2 │
│ Queries Slower                       │         1 │
│ Queries with No Change               │        19 │
│ Queries with Failure                 │         0 │
└──────────────────────────────────────┴───────────┘

Resource Usage

tpch10 — base (merge-base)

Metric	Value
Wall time	42.4s
Peak memory	10.1 GiB
Avg memory	7.8 GiB
CPU user	447.2s
CPU sys	27.7s
Peak spill	0 B

tpch10 — branch

Metric	Value
Wall time	42.0s
Peak memory	10.3 GiB
Avg memory	7.8 GiB
CPU user	441.8s
CPU sys	27.3s
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 perf-convert-to-state
--------------------
Benchmark tpcds_sf1.json
--------------------
┏━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓
┃ Query     ┃                                     HEAD ┃                    perf-convert-to-state ┃        Change ┃
┡━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩
│ QQuery 1  │              6.65 / 7.18 ±0.75 / 8.66 ms │              6.66 / 7.13 ±0.87 / 8.87 ms │     no change │
│ QQuery 2  │        144.93 / 145.96 ±0.81 / 147.32 ms │        143.28 / 144.93 ±0.91 / 145.81 ms │     no change │
│ QQuery 3  │        114.47 / 115.28 ±0.65 / 116.39 ms │        112.50 / 113.16 ±0.56 / 114.15 ms │     no change │
│ QQuery 4  │    1236.40 / 1267.77 ±17.01 / 1281.53 ms │    1231.18 / 1270.32 ±23.40 / 1301.41 ms │     no change │
│ QQuery 5  │        173.62 / 175.39 ±1.37 / 177.07 ms │        171.56 / 173.56 ±1.67 / 175.78 ms │     no change │
│ QQuery 6  │       827.10 / 839.73 ±13.12 / 860.73 ms │       836.29 / 858.54 ±24.90 / 898.93 ms │     no change │
│ QQuery 7  │        342.40 / 345.38 ±1.88 / 347.35 ms │        341.30 / 344.18 ±2.09 / 347.27 ms │     no change │
│ QQuery 8  │        116.61 / 117.36 ±0.65 / 118.21 ms │        114.10 / 115.98 ±1.06 / 116.96 ms │     no change │
│ QQuery 9  │         95.76 / 101.74 ±4.01 / 108.15 ms │        101.40 / 103.40 ±1.54 / 105.90 ms │     no change │
│ QQuery 10 │        104.77 / 106.87 ±1.43 / 109.22 ms │        106.26 / 106.60 ±0.30 / 107.11 ms │     no change │
│ QQuery 11 │        859.38 / 869.26 ±7.67 / 881.14 ms │        859.11 / 865.24 ±5.14 / 874.36 ms │     no change │
│ QQuery 12 │           43.77 / 45.86 ±1.20 / 47.22 ms │           44.31 / 45.13 ±1.17 / 47.43 ms │     no change │
│ QQuery 13 │        400.75 / 404.21 ±2.87 / 407.99 ms │        401.69 / 403.57 ±2.45 / 407.97 ms │     no change │
│ QQuery 14 │      992.52 / 1006.44 ±7.86 / 1016.30 ms │      991.09 / 1005.26 ±9.27 / 1013.24 ms │     no change │
│ QQuery 15 │           15.32 / 17.13 ±1.36 / 19.17 ms │           15.54 / 16.26 ±1.01 / 18.25 ms │ +1.05x faster │
│ QQuery 16 │              7.26 / 7.75 ±0.67 / 9.07 ms │              7.32 / 8.37 ±0.92 / 9.83 ms │  1.08x slower │
│ QQuery 17 │        228.54 / 230.88 ±2.51 / 235.72 ms │        227.05 / 228.97 ±1.33 / 230.75 ms │     no change │
│ QQuery 18 │        126.38 / 128.08 ±1.34 / 130.28 ms │        126.38 / 127.89 ±1.49 / 130.29 ms │     no change │
│ QQuery 19 │        154.02 / 155.90 ±1.43 / 157.62 ms │        154.54 / 155.94 ±0.89 / 157.10 ms │     no change │
│ QQuery 20 │           13.65 / 14.23 ±0.52 / 15.09 ms │           14.33 / 15.01 ±0.50 / 15.75 ms │  1.05x slower │
│ QQuery 21 │           18.78 / 19.45 ±0.56 / 20.18 ms │           18.93 / 19.42 ±0.35 / 19.88 ms │     no change │
│ QQuery 22 │        485.01 / 487.85 ±1.89 / 490.66 ms │        481.71 / 489.37 ±4.00 / 492.95 ms │     no change │
│ QQuery 23 │        841.90 / 857.65 ±9.51 / 871.82 ms │       847.40 / 864.14 ±13.78 / 886.79 ms │     no change │
│ QQuery 24 │        379.02 / 383.71 ±4.19 / 390.82 ms │        380.69 / 384.12 ±2.45 / 388.19 ms │     no change │
│ QQuery 25 │        340.15 / 342.60 ±1.98 / 345.54 ms │        340.87 / 342.62 ±0.92 / 343.60 ms │     no change │
│ QQuery 26 │           82.06 / 83.21 ±1.10 / 84.93 ms │           81.01 / 82.74 ±1.54 / 85.34 ms │     no change │
│ QQuery 27 │              6.77 / 7.20 ±0.51 / 8.20 ms │              6.87 / 7.06 ±0.17 / 7.32 ms │     no change │
│ QQuery 28 │        149.70 / 151.40 ±1.61 / 153.85 ms │        149.13 / 149.97 ±0.70 / 150.86 ms │     no change │
│ QQuery 29 │        280.51 / 282.67 ±1.93 / 285.63 ms │        281.51 / 282.80 ±1.76 / 286.28 ms │     no change │
│ QQuery 30 │           42.63 / 44.28 ±1.07 / 45.38 ms │           44.32 / 45.40 ±0.96 / 46.92 ms │     no change │
│ QQuery 31 │        170.08 / 171.99 ±1.78 / 175.12 ms │        169.73 / 171.74 ±1.86 / 174.21 ms │     no change │
│ QQuery 32 │          56.01 / 65.55 ±17.07 / 99.67 ms │          56.26 / 65.19 ±16.96 / 99.11 ms │     no change │
│ QQuery 33 │        139.33 / 140.29 ±1.03 / 142.22 ms │        139.59 / 140.85 ±1.21 / 142.69 ms │     no change │
│ QQuery 34 │              6.97 / 7.20 ±0.25 / 7.67 ms │              6.99 / 7.53 ±0.64 / 8.65 ms │     no change │
│ QQuery 35 │        104.15 / 106.09 ±1.79 / 109.42 ms │        107.11 / 108.49 ±1.24 / 110.60 ms │     no change │
│ QQuery 36 │              6.49 / 6.77 ±0.31 / 7.38 ms │              6.71 / 6.96 ±0.17 / 7.22 ms │     no change │
│ QQuery 37 │             8.16 / 8.92 ±0.72 / 10.24 ms │              8.19 / 8.84 ±0.41 / 9.31 ms │     no change │
│ QQuery 38 │           84.37 / 88.84 ±4.32 / 96.78 ms │           82.36 / 85.31 ±1.76 / 87.04 ms │     no change │
│ QQuery 39 │        124.61 / 126.62 ±1.31 / 128.11 ms │        125.04 / 127.62 ±1.92 / 130.26 ms │     no change │
│ QQuery 40 │        107.16 / 112.79 ±4.88 / 120.85 ms │        108.92 / 114.30 ±7.93 / 129.96 ms │     no change │
│ QQuery 41 │           14.42 / 14.86 ±0.51 / 15.79 ms │           14.18 / 15.11 ±0.74 / 16.18 ms │     no change │
│ QQuery 42 │        107.10 / 108.76 ±1.44 / 111.08 ms │        105.71 / 108.05 ±1.42 / 109.87 ms │     no change │
│ QQuery 43 │              6.00 / 6.10 ±0.10 / 6.28 ms │              6.10 / 6.66 ±0.85 / 8.33 ms │  1.09x slower │
│ QQuery 44 │           11.49 / 12.17 ±0.64 / 13.38 ms │           11.93 / 12.25 ±0.35 / 12.93 ms │     no change │
│ QQuery 45 │           49.52 / 50.93 ±1.05 / 52.32 ms │           49.81 / 51.55 ±1.44 / 53.96 ms │     no change │
│ QQuery 46 │              8.47 / 8.85 ±0.55 / 9.94 ms │              8.75 / 8.95 ±0.25 / 9.43 ms │     no change │
│ QQuery 47 │        687.14 / 692.02 ±3.23 / 695.31 ms │        684.71 / 690.25 ±4.16 / 697.03 ms │     no change │
│ QQuery 48 │        291.95 / 297.57 ±4.47 / 303.23 ms │        289.55 / 292.23 ±2.14 / 295.53 ms │     no change │
│ QQuery 49 │        252.40 / 256.05 ±2.51 / 259.61 ms │        250.68 / 254.43 ±2.06 / 256.85 ms │     no change │
│ QQuery 50 │        220.87 / 224.87 ±2.63 / 227.78 ms │        219.59 / 225.04 ±3.80 / 228.99 ms │     no change │
│ QQuery 51 │        181.47 / 184.61 ±2.97 / 189.99 ms │        179.77 / 183.58 ±3.30 / 189.30 ms │     no change │
│ QQuery 52 │        107.77 / 109.31 ±1.20 / 110.97 ms │        105.40 / 107.61 ±1.63 / 110.02 ms │     no change │
│ QQuery 53 │        102.83 / 103.72 ±0.51 / 104.21 ms │        100.79 / 101.54 ±0.50 / 102.20 ms │     no change │
│ QQuery 54 │        146.36 / 147.88 ±1.23 / 149.50 ms │        146.08 / 147.44 ±1.49 / 150.15 ms │     no change │
│ QQuery 55 │        106.59 / 109.27 ±1.94 / 111.84 ms │        105.30 / 106.53 ±0.86 / 107.89 ms │     no change │
│ QQuery 56 │        142.41 / 143.44 ±1.01 / 145.01 ms │        138.95 / 140.45 ±1.66 / 143.53 ms │     no change │
│ QQuery 57 │        174.36 / 176.54 ±1.26 / 178.18 ms │        174.56 / 175.17 ±0.46 / 175.98 ms │     no change │
│ QQuery 58 │       286.62 / 303.15 ±13.77 / 326.23 ms │        289.70 / 294.85 ±4.31 / 300.03 ms │     no change │
│ QQuery 59 │        198.00 / 199.48 ±1.15 / 201.52 ms │        196.52 / 198.50 ±1.13 / 199.87 ms │     no change │
│ QQuery 60 │        143.39 / 145.03 ±1.30 / 146.70 ms │        143.82 / 145.74 ±1.51 / 147.68 ms │     no change │
│ QQuery 61 │           12.84 / 13.02 ±0.17 / 13.34 ms │           13.17 / 13.42 ±0.32 / 14.04 ms │     no change │
│ QQuery 62 │      902.55 / 974.74 ±43.01 / 1018.69 ms │      871.14 / 924.65 ±62.42 / 1023.94 ms │ +1.05x faster │
│ QQuery 63 │        101.77 / 107.54 ±7.54 / 122.09 ms │        102.72 / 104.82 ±3.46 / 111.71 ms │     no change │
│ QQuery 64 │        680.27 / 687.39 ±5.53 / 694.81 ms │        685.42 / 686.97 ±1.30 / 688.69 ms │     no change │
│ QQuery 65 │        245.25 / 249.53 ±2.88 / 253.69 ms │        250.06 / 252.27 ±1.35 / 253.85 ms │     no change │
│ QQuery 66 │       245.02 / 258.74 ±12.28 / 276.34 ms │       238.53 / 254.34 ±12.50 / 267.93 ms │     no change │
│ QQuery 67 │        306.82 / 310.24 ±2.64 / 313.10 ms │        303.99 / 309.12 ±3.40 / 314.67 ms │     no change │
│ QQuery 68 │            9.06 / 10.34 ±1.19 / 12.21 ms │            9.18 / 10.94 ±1.79 / 14.23 ms │  1.06x slower │
│ QQuery 69 │        100.65 / 103.15 ±1.60 / 105.40 ms │        100.84 / 103.83 ±3.07 / 108.73 ms │     no change │
│ QQuery 70 │       342.42 / 352.51 ±10.91 / 373.52 ms │        341.91 / 349.02 ±6.98 / 361.50 ms │     no change │
│ QQuery 71 │        134.62 / 137.32 ±1.96 / 140.04 ms │        133.88 / 135.91 ±1.08 / 136.87 ms │     no change │
│ QQuery 72 │       597.51 / 619.78 ±12.82 / 636.84 ms │        607.45 / 616.59 ±6.40 / 623.81 ms │     no change │
│ QQuery 73 │              7.69 / 8.02 ±0.47 / 8.96 ms │              6.58 / 7.82 ±0.84 / 8.62 ms │     no change │
│ QQuery 74 │        528.30 / 538.19 ±5.62 / 545.67 ms │        537.10 / 541.98 ±4.70 / 547.81 ms │     no change │
│ QQuery 75 │        276.32 / 277.50 ±0.94 / 279.12 ms │        274.45 / 277.42 ±1.52 / 278.50 ms │     no change │
│ QQuery 76 │        132.40 / 133.64 ±1.06 / 134.94 ms │        130.83 / 133.55 ±1.95 / 135.90 ms │     no change │
│ QQuery 77 │        186.76 / 188.18 ±1.21 / 190.40 ms │        188.07 / 190.41 ±1.38 / 192.27 ms │     no change │
│ QQuery 78 │        340.00 / 343.67 ±2.95 / 347.74 ms │        338.02 / 341.54 ±2.11 / 343.81 ms │     no change │
│ QQuery 79 │        230.59 / 233.00 ±2.07 / 235.90 ms │        231.73 / 233.61 ±1.18 / 235.00 ms │     no change │
│ QQuery 80 │        321.44 / 324.82 ±3.59 / 331.42 ms │        320.45 / 323.51 ±2.44 / 326.53 ms │     no change │
│ QQuery 81 │           25.89 / 27.41 ±1.31 / 29.84 ms │           25.74 / 27.18 ±1.13 / 28.59 ms │     no change │
│ QQuery 82 │        194.57 / 197.87 ±2.42 / 201.69 ms │        195.33 / 198.35 ±2.06 / 201.23 ms │     no change │
│ QQuery 83 │           37.87 / 38.85 ±0.66 / 39.90 ms │           37.84 / 39.32 ±1.30 / 40.95 ms │     no change │
│ QQuery 84 │           48.28 / 49.31 ±1.48 / 52.22 ms │           47.48 / 48.70 ±1.06 / 50.42 ms │     no change │
│ QQuery 85 │        147.56 / 149.80 ±2.06 / 153.38 ms │        147.79 / 149.55 ±1.32 / 151.05 ms │     no change │
│ QQuery 86 │           38.36 / 39.54 ±0.65 / 40.14 ms │           38.61 / 39.35 ±1.04 / 41.42 ms │     no change │
│ QQuery 87 │           82.13 / 87.76 ±3.75 / 93.85 ms │           85.08 / 88.34 ±2.48 / 92.37 ms │     no change │
│ QQuery 88 │         99.55 / 100.82 ±1.92 / 104.65 ms │         99.48 / 100.41 ±0.73 / 101.66 ms │     no change │
│ QQuery 89 │        116.18 / 118.67 ±1.49 / 120.78 ms │        116.55 / 118.04 ±1.71 / 121.39 ms │     no change │
│ QQuery 90 │           22.58 / 23.14 ±0.38 / 23.54 ms │           23.63 / 24.10 ±0.40 / 24.84 ms │     no change │
│ QQuery 91 │           61.36 / 63.30 ±1.11 / 64.53 ms │           62.23 / 63.33 ±0.78 / 64.30 ms │     no change │
│ QQuery 92 │           56.19 / 57.23 ±0.82 / 58.53 ms │           56.15 / 57.05 ±0.67 / 57.86 ms │     no change │
│ QQuery 93 │        184.34 / 187.30 ±2.25 / 190.40 ms │        186.32 / 186.77 ±0.52 / 187.73 ms │     no change │
│ QQuery 94 │           60.72 / 61.44 ±0.48 / 62.10 ms │           60.38 / 61.60 ±0.75 / 62.57 ms │     no change │
│ QQuery 95 │        128.37 / 129.23 ±1.06 / 131.31 ms │        125.79 / 127.72 ±1.01 / 128.56 ms │     no change │
│ QQuery 96 │           72.03 / 73.37 ±0.80 / 74.15 ms │           73.12 / 74.07 ±0.67 / 75.16 ms │     no change │
│ QQuery 97 │        124.29 / 125.94 ±1.23 / 127.81 ms │        124.90 / 126.77 ±0.94 / 127.43 ms │     no change │
│ QQuery 98 │        155.01 / 156.18 ±1.51 / 159.09 ms │        152.83 / 155.40 ±2.56 / 158.66 ms │     no change │
│ QQuery 99 │ 10689.19 / 10739.27 ±34.07 / 10786.65 ms │ 10675.41 / 10721.08 ±27.97 / 10749.28 ms │     no change │
└───────────┴──────────────────────────────────────────┴──────────────────────────────────────────┴───────────────┘
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━┓
┃ Benchmark Summary                    ┃            ┃
┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━┩
│ Total Time (HEAD)                    │ 31189.86ms │
│ Total Time (perf-convert-to-state)   │ 31098.69ms │
│ Average Time (HEAD)                  │   315.05ms │
│ Average Time (perf-convert-to-state) │   314.13ms │
│ Queries Faster                       │          2 │
│ Queries Slower                       │          4 │
│ Queries with No Change               │         93 │
│ Queries with Failure                 │          0 │
└──────────────────────────────────────┴────────────┘

Resource Usage

tpcds — base (merge-base)

Metric	Value
Wall time	156.3s
Peak memory	5.5 GiB
Avg memory	4.4 GiB
CPU user	257.4s
CPU sys	17.2s
Peak spill	0 B

tpcds — branch

Metric	Value
Wall time	155.8s
Peak memory	5.7 GiB
Avg memory	4.6 GiB
CPU user	257.1s
CPU sys	17.1s
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 perf-convert-to-state
--------------------
Benchmark clickbench_partitioned.json
--------------------
┏━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓
┃ Query     ┃                                  HEAD ┃                 perf-convert-to-state ┃        Change ┃
┡━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩
│ QQuery 0  │          1.32 / 4.68 ±6.52 / 17.72 ms │          1.30 / 4.68 ±6.54 / 17.76 ms │     no change │
│ QQuery 1  │        14.76 / 15.29 ±0.29 / 15.60 ms │        14.64 / 15.45 ±0.59 / 16.42 ms │     no change │
│ QQuery 2  │        44.93 / 45.20 ±0.15 / 45.37 ms │        44.60 / 44.94 ±0.20 / 45.14 ms │     no change │
│ QQuery 3  │        43.80 / 46.98 ±2.01 / 49.27 ms │        42.51 / 42.84 ±0.31 / 43.38 ms │ +1.10x faster │
│ QQuery 4  │     306.15 / 321.69 ±8.22 / 329.41 ms │     321.24 / 323.96 ±2.40 / 327.32 ms │     no change │
│ QQuery 5  │     372.09 / 378.13 ±3.49 / 382.17 ms │     377.72 / 381.71 ±2.52 / 385.33 ms │     no change │
│ QQuery 6  │           6.02 / 7.22 ±1.07 / 9.02 ms │           6.35 / 7.03 ±0.45 / 7.60 ms │     no change │
│ QQuery 7  │        17.23 / 18.17 ±1.02 / 20.09 ms │        17.71 / 18.25 ±0.80 / 19.80 ms │     no change │
│ QQuery 8  │     450.63 / 465.77 ±9.88 / 476.82 ms │    446.44 / 463.57 ±10.60 / 478.21 ms │     no change │
│ QQuery 9  │    711.80 / 731.38 ±10.69 / 741.06 ms │    727.47 / 737.63 ±14.66 / 766.53 ms │     no change │
│ QQuery 10 │      96.84 / 100.04 ±5.29 / 110.59 ms │      98.43 / 100.13 ±1.56 / 102.56 ms │     no change │
│ QQuery 11 │     108.70 / 111.51 ±3.15 / 116.77 ms │     111.47 / 112.32 ±0.92 / 113.98 ms │     no change │
│ QQuery 12 │     373.21 / 378.10 ±4.07 / 385.13 ms │     377.43 / 385.10 ±5.98 / 392.88 ms │     no change │
│ QQuery 13 │    499.45 / 520.51 ±20.31 / 552.60 ms │     498.92 / 514.70 ±8.65 / 522.11 ms │     no change │
│ QQuery 14 │     363.08 / 373.85 ±6.39 / 381.56 ms │     382.67 / 387.16 ±4.43 / 393.13 ms │     no change │
│ QQuery 15 │     390.98 / 399.68 ±7.75 / 410.96 ms │     398.91 / 409.61 ±8.23 / 421.47 ms │     no change │
│ QQuery 16 │     763.35 / 779.87 ±9.15 / 787.41 ms │     780.65 / 789.63 ±8.07 / 799.39 ms │     no change │
│ QQuery 17 │     761.35 / 768.55 ±4.90 / 776.74 ms │     770.70 / 781.22 ±5.56 / 787.18 ms │     no change │
│ QQuery 18 │ 1450.98 / 1492.13 ±37.73 / 1562.61 ms │ 1552.89 / 1622.25 ±42.65 / 1671.75 ms │  1.09x slower │
│ QQuery 19 │      46.11 / 71.50 ±25.61 / 107.84 ms │        37.72 / 40.35 ±2.17 / 44.00 ms │ +1.77x faster │
│ QQuery 20 │    740.83 / 767.08 ±26.42 / 812.66 ms │    745.57 / 773.43 ±23.30 / 812.19 ms │     no change │
│ QQuery 21 │     789.77 / 794.02 ±3.51 / 798.83 ms │     792.69 / 800.03 ±7.53 / 814.55 ms │     no change │
│ QQuery 22 │  1177.43 / 1185.35 ±7.01 / 1198.44 ms │  1196.53 / 1206.13 ±6.47 / 1215.05 ms │     no change │
│ QQuery 23 │ 3222.26 / 3250.58 ±36.44 / 3318.28 ms │ 3277.70 / 3299.06 ±18.87 / 3326.66 ms │     no change │
│ QQuery 24 │     100.81 / 109.43 ±5.61 / 116.24 ms │     107.20 / 111.05 ±2.80 / 115.06 ms │     no change │
│ QQuery 25 │     142.64 / 144.62 ±1.72 / 147.76 ms │     144.72 / 147.41 ±2.23 / 150.39 ms │     no change │
│ QQuery 26 │     104.92 / 108.09 ±2.53 / 110.72 ms │     106.88 / 108.61 ±0.92 / 109.59 ms │     no change │
│ QQuery 27 │     876.35 / 883.10 ±8.12 / 895.75 ms │    886.21 / 892.33 ±10.16 / 912.56 ms │     no change │
│ QQuery 28 │ 7809.62 / 7888.01 ±48.46 / 7944.16 ms │ 7854.64 / 7915.39 ±36.17 / 7967.20 ms │     no change │
│ QQuery 29 │        53.27 / 57.22 ±4.90 / 66.39 ms │        53.97 / 60.55 ±7.22 / 74.40 ms │  1.06x slower │
│ QQuery 30 │     394.86 / 400.95 ±6.07 / 410.33 ms │     400.19 / 403.22 ±2.98 / 408.17 ms │     no change │
│ QQuery 31 │     405.76 / 421.10 ±8.15 / 428.34 ms │    397.03 / 416.49 ±15.88 / 435.18 ms │     no change │
│ QQuery 32 │ 1149.12 / 1162.67 ±15.24 / 1190.80 ms │ 1119.73 / 1155.16 ±24.04 / 1194.78 ms │     no change │
│ QQuery 33 │ 1619.52 / 1634.06 ±11.39 / 1647.80 ms │  1624.41 / 1633.67 ±7.42 / 1644.10 ms │     no change │
│ QQuery 34 │ 1626.35 / 1649.99 ±14.43 / 1668.31 ms │ 1634.91 / 1665.53 ±22.38 / 1700.67 ms │     no change │
│ QQuery 35 │     456.90 / 466.54 ±7.48 / 476.48 ms │     452.52 / 456.35 ±2.37 / 459.09 ms │     no change │
│ QQuery 36 │     121.11 / 127.39 ±3.66 / 130.25 ms │     128.70 / 130.67 ±1.40 / 132.23 ms │     no change │
│ QQuery 37 │        50.74 / 53.65 ±2.32 / 56.95 ms │        50.23 / 52.58 ±1.59 / 54.92 ms │     no change │
│ QQuery 38 │        77.55 / 79.68 ±1.13 / 80.79 ms │        81.00 / 82.03 ±0.71 / 82.85 ms │     no change │
│ QQuery 39 │    230.66 / 239.95 ±10.16 / 253.01 ms │     232.31 / 239.63 ±5.81 / 244.44 ms │     no change │
│ QQuery 40 │        25.69 / 27.71 ±1.31 / 29.18 ms │        25.54 / 27.87 ±1.92 / 31.07 ms │     no change │
│ QQuery 41 │        21.44 / 22.25 ±0.69 / 23.22 ms │        22.10 / 23.95 ±2.05 / 27.87 ms │  1.08x slower │
│ QQuery 42 │        21.46 / 21.97 ±0.29 / 22.21 ms │        21.08 / 21.81 ±0.59 / 22.59 ms │     no change │
└───────────┴───────────────────────────────────────┴───────────────────────────────────────┴───────────────┘
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━┓
┃ Benchmark Summary                    ┃            ┃
┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━┩
│ Total Time (HEAD)                    │ 28525.63ms │
│ Total Time (perf-convert-to-state)   │ 28805.50ms │
│ Average Time (HEAD)                  │   663.39ms │
│ Average Time (perf-convert-to-state) │   669.90ms │
│ Queries Faster                       │          2 │
│ Queries Slower                       │          3 │
│ Queries with No Change               │         38 │
│ Queries with Failure                 │          0 │
└──────────────────────────────────────┴────────────┘

Resource Usage

clickbench_partitioned — base (merge-base)

Metric	Value
Wall time	143.8s
Peak memory	39.9 GiB
Avg memory	30.0 GiB
CPU user	1366.5s
CPU sys	93.0s
Peak spill	0 B

clickbench_partitioned — branch

Metric	Value
Wall time	145.1s
Peak memory	39.0 GiB
Avg memory	27.5 GiB
CPU user	1372.3s
CPU sys	101.3s
Peak spill	0 B

---

File an issue against this benchmark runner