feat: estimate cardinality for semi and anti-joins using distinct counts

[buraksenn]

Which issue does this PR close?

Does not close but part of #20766

Rationale for this change

Details are in #20766. But main idea is to use existing distinct count information to optimize joins similar to how Spark/Trino does

What changes are included in this PR?

This PR extends cardinality estimation for semi/anti joins using distinct counts

Are these changes tested?

I've added cases but not sure if I should've added benchmarks on this.

Are there any user-facing changes?

No

[asolimando]

LGTM, a couple of minor points and a few tests to be added. The only change I'd like to see is bailing out when either side has no stats for a column pair.

[asolimando]

LGTM, thanks for addressing all my comments fully!

[Dandandan]

Can we run some benchmarks (e.g. tpch, tpcds, ...) for PRs like these in the future?

Would be good to see impact of individual contributions and possibly catch regressions / issues.

[xudong963]

run benchmarks

[xudong963]

Can we run some benchmarks (e.g. tpch, tpcds, ...) for PRs like these in the future?

Would be good to see impact of individual contributions and possibly catch regressions / issues.

Sure, running

[xudong963]

run benchmarks tpcds tpch

env:
  DATAFUSION_EXECUTION_PARQUET_PUSHDOWN_FILTERS: true
  DATAFUSION_EXECUTION_PARQUET_REORDER_FILTERS: true

[adriangbot]

🤖 Benchmark running (GKE) | trigger
Instance: c4a-highmem-16 (12 vCPU / 65 GiB) | Linux bench-c4286657778-1656-zzhxn 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 use-ndv-for-semi-and-anti-join (5978b00) to 7acbe03 (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-c4286657778-1657-pcv4l 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 use-ndv-for-semi-and-anti-join (5978b00) to 7acbe03 (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-c4286664748-1659-x7rvl 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 use-ndv-for-semi-and-anti-join (5978b00) to 7acbe03 (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-c4286664748-1660-jltq5 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 use-ndv-for-semi-and-anti-join (5978b00) to 7acbe03 (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-c4286657778-1658-77bg4 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 use-ndv-for-semi-and-anti-join (5978b00) to 7acbe03 (merge-base) diff using: tpch
Results will be posted here when complete

---

File an issue against this benchmark runner

[Dandandan]

run benchmarks tpcds tpch

[adriangbot]

🤖 Benchmark running (GKE) | trigger
Instance: c4a-highmem-16 (12 vCPU / 65 GiB) | Linux bench-c4286764315-1667-vb9vk 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 use-ndv-for-semi-and-anti-join (5978b00) to 7acbe03 (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-c4286764315-1668-84wt4 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 use-ndv-for-semi-and-anti-join (5978b00) to 7acbe03 (merge-base) diff using: tpch
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 use-ndv-for-semi-and-anti-join
--------------------
Benchmark clickbench_partitioned.json
--------------------
┏━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓
┃ Query     ┃                                  HEAD ┃        use-ndv-for-semi-and-anti-join ┃        Change ┃
┡━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩
│ QQuery 0  │          1.16 / 4.39 ±6.33 / 17.05 ms │          1.17 / 4.35 ±6.27 / 16.90 ms │     no change │
│ QQuery 1  │        12.42 / 13.01 ±0.36 / 13.52 ms │        12.35 / 12.88 ±0.30 / 13.28 ms │     no change │
│ QQuery 2  │        36.92 / 37.06 ±0.20 / 37.46 ms │        36.45 / 36.58 ±0.12 / 36.75 ms │     no change │
│ QQuery 3  │        31.19 / 31.97 ±0.81 / 33.47 ms │        31.50 / 31.66 ±0.14 / 31.88 ms │     no change │
│ QQuery 4  │     253.24 / 255.57 ±2.48 / 259.20 ms │     247.31 / 260.80 ±8.16 / 272.66 ms │     no change │
│ QQuery 5  │     294.44 / 297.06 ±2.11 / 299.49 ms │     293.68 / 302.22 ±5.36 / 310.44 ms │     no change │
│ QQuery 6  │           5.62 / 7.17 ±1.17 / 8.98 ms │           6.29 / 7.32 ±0.60 / 8.17 ms │     no change │
│ QQuery 7  │        14.46 / 14.69 ±0.13 / 14.86 ms │        13.63 / 13.95 ±0.35 / 14.61 ms │ +1.05x faster │
│ QQuery 8  │     362.09 / 369.96 ±6.49 / 378.06 ms │     368.28 / 372.45 ±4.79 / 380.07 ms │     no change │
│ QQuery 9  │     532.66 / 538.56 ±5.16 / 546.06 ms │     511.32 / 520.28 ±8.85 / 536.40 ms │     no change │
│ QQuery 10 │        76.20 / 77.74 ±1.84 / 81.00 ms │        74.38 / 76.88 ±2.24 / 81.08 ms │     no change │
│ QQuery 11 │        85.67 / 86.13 ±0.36 / 86.61 ms │        85.57 / 85.88 ±0.36 / 86.53 ms │     no change │
│ QQuery 12 │     291.12 / 298.04 ±4.58 / 303.68 ms │    294.57 / 308.24 ±10.77 / 326.16 ms │     no change │
│ QQuery 13 │    425.30 / 450.48 ±30.14 / 507.68 ms │    424.82 / 443.19 ±19.72 / 475.27 ms │     no change │
│ QQuery 14 │     305.45 / 310.84 ±5.25 / 320.60 ms │     309.30 / 312.95 ±2.59 / 316.51 ms │     no change │
│ QQuery 15 │     322.75 / 332.62 ±7.53 / 345.19 ms │    328.94 / 347.13 ±19.35 / 384.27 ms │     no change │
│ QQuery 16 │    686.53 / 707.14 ±16.50 / 726.29 ms │     683.33 / 689.75 ±5.43 / 699.17 ms │     no change │
│ QQuery 17 │     678.14 / 684.44 ±4.80 / 691.75 ms │     681.96 / 691.09 ±6.14 / 700.38 ms │     no change │
│ QQuery 18 │ 1429.11 / 1481.99 ±42.96 / 1556.50 ms │ 1350.51 / 1366.00 ±19.79 / 1403.41 ms │ +1.08x faster │
│ QQuery 19 │        30.45 / 35.12 ±8.91 / 52.94 ms │        30.59 / 31.61 ±1.05 / 33.26 ms │ +1.11x faster │
│ QQuery 20 │    518.45 / 531.27 ±12.94 / 548.44 ms │     512.68 / 518.49 ±6.08 / 529.71 ms │     no change │
│ QQuery 21 │     592.25 / 600.87 ±8.70 / 616.07 ms │     595.75 / 598.07 ±1.55 / 599.98 ms │     no change │
│ QQuery 22 │  1055.13 / 1062.36 ±4.71 / 1067.62 ms │  1055.87 / 1057.49 ±1.33 / 1059.37 ms │     no change │
│ QQuery 23 │ 3291.59 / 3315.49 ±23.15 / 3357.33 ms │  3282.80 / 3294.55 ±8.26 / 3303.69 ms │     no change │
│ QQuery 24 │        45.77 / 47.69 ±2.77 / 53.16 ms │        43.38 / 45.27 ±2.20 / 49.39 ms │ +1.05x faster │
│ QQuery 25 │     115.21 / 116.55 ±1.14 / 117.93 ms │     116.11 / 117.36 ±0.64 / 117.91 ms │     no change │
│ QQuery 26 │        42.96 / 44.38 ±1.15 / 46.37 ms │        42.50 / 43.39 ±0.62 / 44.01 ms │     no change │
│ QQuery 27 │     663.10 / 667.60 ±4.25 / 675.70 ms │     665.76 / 670.11 ±5.11 / 679.14 ms │     no change │
│ QQuery 28 │ 2979.09 / 3023.24 ±30.17 / 3072.57 ms │ 2975.67 / 3021.18 ±24.62 / 3050.60 ms │     no change │
│ QQuery 29 │        44.36 / 47.96 ±3.00 / 52.56 ms │        43.89 / 48.66 ±3.92 / 55.01 ms │     no change │
│ QQuery 30 │     327.66 / 332.63 ±5.01 / 341.76 ms │     331.13 / 336.39 ±6.13 / 347.70 ms │     no change │
│ QQuery 31 │     337.39 / 345.20 ±5.45 / 354.18 ms │     326.73 / 330.91 ±6.02 / 342.59 ms │     no change │
│ QQuery 32 │ 1156.76 / 1239.36 ±52.89 / 1322.58 ms │  1018.91 / 1027.33 ±6.66 / 1038.90 ms │ +1.21x faster │
│ QQuery 33 │ 1478.22 / 1509.01 ±31.72 / 1560.59 ms │ 1437.45 / 1463.57 ±25.85 / 1497.89 ms │     no change │
│ QQuery 34 │ 1498.76 / 1594.43 ±65.90 / 1678.40 ms │ 1469.84 / 1486.05 ±19.68 / 1524.64 ms │ +1.07x faster │
│ QQuery 35 │     321.71 / 327.91 ±6.85 / 341.08 ms │     320.46 / 325.82 ±6.52 / 338.19 ms │     no change │
│ QQuery 36 │        61.59 / 65.13 ±3.26 / 70.84 ms │        62.44 / 67.38 ±2.82 / 70.42 ms │     no change │
│ QQuery 37 │        37.47 / 39.20 ±1.18 / 40.54 ms │        37.45 / 39.22 ±1.30 / 41.44 ms │     no change │
│ QQuery 38 │        41.04 / 42.41 ±0.93 / 43.61 ms │        40.30 / 42.37 ±1.43 / 44.12 ms │     no change │
│ QQuery 39 │     125.53 / 130.77 ±5.20 / 139.78 ms │     119.41 / 128.71 ±5.86 / 137.25 ms │     no change │
│ QQuery 40 │        14.61 / 17.44 ±2.33 / 19.90 ms │        14.76 / 16.33 ±1.40 / 18.35 ms │ +1.07x faster │
│ QQuery 41 │        15.19 / 16.12 ±1.11 / 18.16 ms │        13.59 / 14.80 ±1.25 / 16.86 ms │ +1.09x faster │
│ QQuery 42 │        13.62 / 14.11 ±0.38 / 14.75 ms │        13.33 / 14.54 ±1.08 / 16.02 ms │     no change │
└───────────┴───────────────────────────────────────┴───────────────────────────────────────┴───────────────┘
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━┓
┃ Benchmark Summary                             ┃            ┃
┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━┩
│ Total Time (HEAD)                             │ 21165.10ms │
│ Total Time (use-ndv-for-semi-and-anti-join)   │ 20623.20ms │
│ Average Time (HEAD)                           │   492.21ms │
│ Average Time (use-ndv-for-semi-and-anti-join) │   479.61ms │
│ Queries Faster                                │          8 │
│ Queries Slower                                │          0 │
│ Queries with No Change                        │         35 │
│ Queries with Failure                          │          0 │
└───────────────────────────────────────────────┴────────────┘

Resource Usage

clickbench_partitioned — base (merge-base)

Metric	Value
Wall time	110.0s
Peak memory	34.9 GiB
Avg memory	27.1 GiB
CPU user	1082.5s
CPU sys	103.6s
Peak spill	0 B

clickbench_partitioned — branch

Metric	Value
Wall time	105.0s
Peak memory	39.4 GiB
Avg memory	33.5 GiB
CPU user	1079.5s
CPU sys	84.3s
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 use-ndv-for-semi-and-anti-join
--------------------
Benchmark tpcds_sf1.json
--------------------
┏━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━┓
┃ Query     ┃                                     HEAD ┃           use-ndv-for-semi-and-anti-join ┃         Change ┃
┡━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━┩
│ QQuery 1  │              6.42 / 6.78 ±0.67 / 8.11 ms │              6.42 / 6.82 ±0.71 / 8.23 ms │      no change │
│ QQuery 2  │        145.82 / 148.54 ±3.64 / 155.69 ms │        146.56 / 147.19 ±0.56 / 148.21 ms │      no change │
│ QQuery 3  │        113.69 / 114.01 ±0.35 / 114.64 ms │        113.62 / 114.72 ±1.57 / 117.78 ms │      no change │
│ QQuery 4  │    1286.53 / 1338.12 ±55.39 / 1445.73 ms │   1254.43 / 1325.91 ±104.15 / 1530.00 ms │      no change │
│ QQuery 5  │        176.74 / 178.24 ±1.56 / 180.85 ms │        178.65 / 179.58 ±0.63 / 180.30 ms │      no change │
│ QQuery 6  │       878.14 / 904.35 ±26.38 / 943.08 ms │       834.43 / 897.42 ±35.56 / 930.46 ms │      no change │
│ QQuery 7  │        341.84 / 344.52 ±4.08 / 352.60 ms │        343.18 / 344.75 ±1.43 / 347.14 ms │      no change │
│ QQuery 8  │        115.65 / 117.21 ±1.02 / 118.35 ms │        115.51 / 116.94 ±1.04 / 118.46 ms │      no change │
│ QQuery 9  │        100.01 / 103.03 ±2.46 / 106.95 ms │        100.06 / 103.47 ±5.49 / 114.39 ms │      no change │
│ QQuery 10 │        104.99 / 106.30 ±1.10 / 108.20 ms │        104.67 / 106.15 ±1.06 / 107.16 ms │      no change │
│ QQuery 11 │     907.29 / 1041.18 ±69.13 / 1097.53 ms │       909.54 / 932.60 ±12.87 / 946.28 ms │  +1.12x faster │
│ QQuery 12 │           47.43 / 49.97 ±1.96 / 52.48 ms │           44.58 / 47.19 ±2.49 / 51.79 ms │  +1.06x faster │
│ QQuery 13 │       404.77 / 418.39 ±10.79 / 428.45 ms │        396.73 / 402.90 ±4.18 / 407.52 ms │      no change │
│ QQuery 14 │     992.80 / 1009.60 ±11.15 / 1025.37 ms │     998.10 / 1018.35 ±12.92 / 1032.43 ms │      no change │
│ QQuery 15 │           14.92 / 15.67 ±0.66 / 16.70 ms │           17.16 / 18.46 ±0.89 / 19.50 ms │   1.18x slower │
│ QQuery 16 │              6.96 / 7.16 ±0.17 / 7.44 ms │              7.88 / 8.12 ±0.29 / 8.61 ms │   1.13x slower │
│ QQuery 17 │        226.42 / 228.78 ±1.91 / 232.22 ms │        239.05 / 241.89 ±2.52 / 245.43 ms │   1.06x slower │
│ QQuery 18 │        128.50 / 130.98 ±1.92 / 133.29 ms │        132.37 / 133.65 ±1.01 / 135.05 ms │      no change │
│ QQuery 19 │        162.40 / 165.34 ±1.79 / 167.75 ms │        164.26 / 165.44 ±1.27 / 167.62 ms │      no change │
│ QQuery 20 │           14.34 / 14.87 ±0.43 / 15.38 ms │           15.02 / 15.46 ±0.38 / 16.05 ms │      no change │
│ QQuery 21 │           20.64 / 21.03 ±0.27 / 21.37 ms │           20.54 / 21.39 ±0.53 / 22.07 ms │      no change │
│ QQuery 22 │        522.64 / 528.24 ±5.11 / 534.44 ms │       484.35 / 518.11 ±28.09 / 564.95 ms │      no change │
│ QQuery 23 │       853.16 / 871.31 ±10.42 / 883.07 ms │       856.26 / 878.47 ±31.07 / 939.61 ms │      no change │
│ QQuery 24 │        377.48 / 380.17 ±1.82 / 382.26 ms │        405.69 / 409.10 ±3.62 / 415.88 ms │   1.08x slower │
│ QQuery 25 │        338.31 / 340.71 ±1.40 / 342.24 ms │        356.35 / 357.93 ±0.85 / 358.73 ms │   1.05x slower │
│ QQuery 26 │           79.54 / 81.50 ±1.63 / 84.27 ms │           86.08 / 87.54 ±0.97 / 89.10 ms │   1.07x slower │
│ QQuery 27 │              6.63 / 7.05 ±0.53 / 8.07 ms │              7.37 / 7.71 ±0.24 / 8.02 ms │   1.09x slower │
│ QQuery 28 │        147.79 / 149.68 ±2.23 / 153.94 ms │        153.95 / 155.46 ±1.48 / 158.09 ms │      no change │
│ QQuery 29 │        279.63 / 282.07 ±1.98 / 285.04 ms │        281.77 / 290.57 ±5.02 / 295.50 ms │      no change │
│ QQuery 30 │           43.57 / 45.69 ±1.91 / 48.91 ms │           43.58 / 45.33 ±1.23 / 46.78 ms │      no change │
│ QQuery 31 │        178.01 / 179.22 ±1.02 / 180.53 ms │        170.76 / 173.19 ±1.51 / 175.24 ms │      no change │
│ QQuery 32 │           14.27 / 14.47 ±0.14 / 14.71 ms │          13.20 / 22.58 ±17.00 / 56.56 ms │   1.56x slower │
│ QQuery 33 │        147.48 / 150.86 ±2.41 / 153.59 ms │        139.55 / 141.69 ±1.20 / 143.20 ms │  +1.06x faster │
│ QQuery 34 │              7.66 / 7.77 ±0.08 / 7.90 ms │              6.84 / 7.02 ±0.23 / 7.45 ms │  +1.11x faster │
│ QQuery 35 │        112.80 / 116.42 ±2.00 / 118.33 ms │        106.28 / 108.27 ±1.32 / 110.15 ms │  +1.08x faster │
│ QQuery 36 │              6.97 / 7.33 ±0.22 / 7.62 ms │              6.37 / 6.53 ±0.18 / 6.76 ms │  +1.12x faster │
│ QQuery 37 │            9.43 / 10.23 ±0.83 / 11.82 ms │             8.70 / 9.33 ±0.71 / 10.67 ms │  +1.10x faster │
│ QQuery 38 │           89.05 / 93.22 ±2.64 / 96.80 ms │          95.53 / 98.85 ±3.72 / 106.06 ms │   1.06x slower │
│ QQuery 39 │        140.51 / 143.36 ±3.79 / 150.87 ms │        121.73 / 125.61 ±2.78 / 129.53 ms │  +1.14x faster │
│ QQuery 40 │        117.06 / 123.14 ±6.35 / 133.25 ms │        108.48 / 114.59 ±7.58 / 129.13 ms │  +1.07x faster │
│ QQuery 41 │           15.41 / 16.82 ±1.03 / 18.17 ms │           14.26 / 14.92 ±0.44 / 15.42 ms │  +1.13x faster │
│ QQuery 42 │        110.26 / 111.65 ±1.43 / 114.31 ms │        106.45 / 108.38 ±1.09 / 109.83 ms │      no change │
│ QQuery 43 │              5.54 / 5.67 ±0.17 / 6.01 ms │              5.49 / 5.60 ±0.11 / 5.81 ms │      no change │
│ QQuery 44 │           11.28 / 11.97 ±0.69 / 13.29 ms │           11.28 / 11.62 ±0.20 / 11.82 ms │      no change │
│ QQuery 45 │           52.11 / 53.38 ±0.92 / 54.92 ms │           50.71 / 51.52 ±0.65 / 52.57 ms │      no change │
│ QQuery 46 │              7.98 / 8.78 ±0.59 / 9.70 ms │              8.31 / 8.86 ±0.61 / 9.80 ms │      no change │
│ QQuery 47 │       691.19 / 706.01 ±11.59 / 719.46 ms │       698.11 / 792.16 ±68.89 / 860.53 ms │   1.12x slower │
│ QQuery 48 │        286.04 / 289.26 ±1.98 / 291.71 ms │        297.82 / 307.43 ±6.43 / 316.32 ms │   1.06x slower │
│ QQuery 49 │        252.48 / 255.18 ±2.33 / 258.74 ms │        252.18 / 253.87 ±1.93 / 257.55 ms │      no change │
│ QQuery 50 │        226.69 / 229.52 ±3.16 / 235.52 ms │        224.81 / 229.74 ±2.66 / 231.88 ms │      no change │
│ QQuery 51 │        178.17 / 182.91 ±3.28 / 187.05 ms │        180.78 / 184.34 ±3.14 / 188.99 ms │      no change │
│ QQuery 52 │        108.46 / 110.18 ±1.29 / 112.36 ms │        107.22 / 108.58 ±1.10 / 110.29 ms │      no change │
│ QQuery 53 │        102.46 / 104.31 ±1.78 / 107.21 ms │        101.38 / 102.83 ±1.39 / 104.80 ms │      no change │
│ QQuery 54 │        145.90 / 147.87 ±1.49 / 149.86 ms │        145.40 / 145.98 ±0.74 / 147.43 ms │      no change │
│ QQuery 55 │        107.82 / 109.41 ±1.17 / 110.83 ms │        105.84 / 108.80 ±2.22 / 112.16 ms │      no change │
│ QQuery 56 │        139.59 / 141.83 ±1.45 / 143.68 ms │        140.03 / 141.70 ±1.33 / 143.78 ms │      no change │
│ QQuery 57 │        172.93 / 175.35 ±2.09 / 178.16 ms │        169.39 / 172.46 ±1.58 / 173.87 ms │      no change │
│ QQuery 58 │        270.33 / 273.50 ±2.09 / 275.75 ms │        316.52 / 318.68 ±2.35 / 322.84 ms │   1.17x slower │
│ QQuery 59 │        196.83 / 197.93 ±0.79 / 199.30 ms │        196.03 / 198.39 ±2.08 / 200.83 ms │      no change │
│ QQuery 60 │        145.37 / 146.94 ±1.40 / 148.77 ms │        143.24 / 147.51 ±2.64 / 150.75 ms │      no change │
│ QQuery 61 │           13.23 / 13.47 ±0.15 / 13.67 ms │           13.39 / 14.05 ±0.49 / 14.84 ms │      no change │
│ QQuery 62 │      956.55 / 987.26 ±21.23 / 1022.58 ms │     898.28 / 1006.53 ±76.76 / 1127.00 ms │      no change │
│ QQuery 63 │        105.45 / 108.14 ±2.54 / 112.74 ms │        103.15 / 104.81 ±2.52 / 109.80 ms │      no change │
│ QQuery 64 │       676.41 / 689.99 ±13.90 / 716.40 ms │        680.12 / 682.47 ±2.63 / 687.52 ms │      no change │
│ QQuery 65 │        252.50 / 253.95 ±1.41 / 256.54 ms │        251.79 / 253.96 ±1.90 / 257.06 ms │      no change │
│ QQuery 66 │       236.56 / 252.53 ±15.55 / 272.07 ms │       241.15 / 255.27 ±11.41 / 269.74 ms │      no change │
│ QQuery 67 │        319.18 / 332.57 ±6.99 / 338.45 ms │        307.02 / 313.38 ±6.27 / 324.50 ms │  +1.06x faster │
│ QQuery 68 │           10.20 / 11.13 ±1.21 / 13.52 ms │             8.11 / 9.77 ±1.65 / 12.77 ms │  +1.14x faster │
│ QQuery 69 │        105.05 / 107.96 ±4.12 / 116.02 ms │        102.59 / 104.19 ±0.88 / 105.24 ms │      no change │
│ QQuery 70 │       353.73 / 365.09 ±10.24 / 382.23 ms │        341.31 / 347.28 ±6.16 / 358.87 ms │      no change │
│ QQuery 71 │        138.72 / 142.80 ±2.99 / 147.05 ms │        135.56 / 137.19 ±1.34 / 139.31 ms │      no change │
│ QQuery 72 │       654.16 / 670.30 ±14.67 / 689.22 ms │        618.93 / 625.35 ±5.35 / 632.03 ms │  +1.07x faster │
│ QQuery 73 │              7.14 / 8.01 ±0.61 / 8.93 ms │              6.84 / 7.91 ±0.99 / 9.54 ms │      no change │
│ QQuery 74 │        586.86 / 590.46 ±3.13 / 596.21 ms │       583.96 / 654.43 ±42.25 / 692.22 ms │   1.11x slower │
│ QQuery 75 │        276.01 / 278.91 ±1.83 / 281.52 ms │        284.21 / 287.23 ±2.90 / 292.15 ms │      no change │
│ QQuery 76 │        135.73 / 137.29 ±1.15 / 138.47 ms │        137.05 / 138.25 ±1.17 / 140.27 ms │      no change │
│ QQuery 77 │        192.64 / 196.03 ±2.15 / 198.85 ms │        187.24 / 192.29 ±3.69 / 198.32 ms │      no change │
│ QQuery 78 │        355.97 / 364.04 ±4.61 / 369.81 ms │        336.20 / 348.01 ±9.02 / 360.29 ms │      no change │
│ QQuery 79 │        232.58 / 238.11 ±3.46 / 243.02 ms │        230.76 / 232.73 ±1.46 / 234.65 ms │      no change │
│ QQuery 80 │        323.13 / 327.78 ±3.45 / 332.45 ms │        324.11 / 325.71 ±2.23 / 330.09 ms │      no change │
│ QQuery 81 │           25.90 / 27.46 ±1.22 / 29.22 ms │           25.23 / 26.03 ±0.47 / 26.43 ms │  +1.05x faster │
│ QQuery 82 │           40.57 / 41.83 ±1.40 / 44.42 ms │           40.00 / 40.61 ±0.35 / 40.98 ms │      no change │
│ QQuery 83 │           38.41 / 40.52 ±1.98 / 43.66 ms │           37.92 / 39.03 ±0.95 / 40.51 ms │      no change │
│ QQuery 84 │           48.12 / 48.83 ±0.55 / 49.53 ms │           47.56 / 49.53 ±1.84 / 52.15 ms │      no change │
│ QQuery 85 │        145.95 / 147.53 ±1.57 / 150.17 ms │        146.54 / 148.18 ±1.12 / 149.67 ms │      no change │
│ QQuery 86 │           37.28 / 39.18 ±1.11 / 40.59 ms │           40.55 / 41.33 ±0.62 / 42.42 ms │   1.05x slower │
│ QQuery 87 │           85.75 / 88.07 ±2.62 / 93.17 ms │              3.60 / 3.80 ±0.19 / 4.09 ms │ +23.19x faster │
│ QQuery 88 │           98.00 / 99.06 ±0.62 / 99.89 ms │        103.22 / 103.76 ±0.37 / 104.30 ms │      no change │
│ QQuery 89 │        119.44 / 120.19 ±0.54 / 120.77 ms │        121.17 / 123.76 ±1.45 / 125.15 ms │      no change │
│ QQuery 90 │           23.06 / 23.50 ±0.38 / 24.02 ms │           24.83 / 25.46 ±0.34 / 25.75 ms │   1.08x slower │
│ QQuery 91 │           61.38 / 63.31 ±1.06 / 64.56 ms │           66.99 / 68.25 ±0.73 / 69.11 ms │   1.08x slower │
│ QQuery 92 │           56.21 / 57.29 ±1.00 / 59.01 ms │           60.34 / 61.32 ±0.62 / 62.24 ms │   1.07x slower │
│ QQuery 93 │        184.85 / 186.65 ±1.18 / 188.01 ms │        198.84 / 200.36 ±0.99 / 201.80 ms │   1.07x slower │
│ QQuery 94 │           62.25 / 62.59 ±0.39 / 63.11 ms │           65.04 / 67.04 ±1.44 / 69.50 ms │   1.07x slower │
│ QQuery 95 │        129.34 / 129.95 ±0.59 / 130.85 ms │        135.11 / 136.38 ±1.04 / 138.08 ms │      no change │
│ QQuery 96 │           74.15 / 74.77 ±0.57 / 75.69 ms │           77.42 / 78.52 ±1.00 / 80.14 ms │   1.05x slower │
│ QQuery 97 │        124.80 / 125.61 ±0.56 / 126.20 ms │        131.70 / 135.68 ±2.28 / 138.06 ms │   1.08x slower │
│ QQuery 98 │        152.95 / 154.34 ±1.04 / 155.76 ms │        165.03 / 166.91 ±1.38 / 169.11 ms │   1.08x slower │
│ QQuery 99 │ 10725.87 / 10871.89 ±78.17 / 10943.25 ms │ 10706.78 / 10824.92 ±80.61 / 10935.60 ms │      no change │
└───────────┴──────────────────────────────────────────┴──────────────────────────────────────────┴────────────────┘
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━┓
┃ Benchmark Summary                             ┃            ┃
┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━┩
│ Total Time (HEAD)                             │ 31753.39ms │
│ Total Time (use-ndv-for-semi-and-anti-join)   │ 31701.36ms │
│ Average Time (HEAD)                           │   320.74ms │
│ Average Time (use-ndv-for-semi-and-anti-join) │   320.22ms │
│ Queries Faster                                │         15 │
│ Queries Slower                                │         22 │
│ Queries with No Change                        │         62 │
│ Queries with Failure                          │          0 │
└───────────────────────────────────────────────┴────────────┘

Resource Usage

tpcds — base (merge-base)

Metric	Value
Wall time	160.0s
Peak memory	6.7 GiB
Avg memory	5.6 GiB
CPU user	263.5s
CPU sys	17.1s
Peak spill	0 B

tpcds — branch

Metric	Value
Wall time	160.0s
Peak memory	6.8 GiB
Avg memory	5.6 GiB
CPU user	260.6s
CPU sys	17.1s
Peak spill	0 B

---

File an issue against this benchmark runner

[Dandandan]

│ QQuery 87 │ 85.75 / 88.07 ±2.62 / 93.17 ms │ 3.60 / 3.80 ±0.19 / 4.09 ms │ +23.19x faster │

Interesting!

[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 use-ndv-for-semi-and-anti-join
--------------------
Benchmark tpcds_sf1.json
--------------------
┏━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━┓
┃ Query     ┃                                      HEAD ┃            use-ndv-for-semi-and-anti-join ┃         Change ┃
┡━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━┩
│ QQuery 1  │               6.15 / 6.55 ±0.72 / 7.99 ms │               6.17 / 6.64 ±0.81 / 8.25 ms │      no change │
│ QQuery 2  │         112.57 / 113.13 ±0.71 / 114.50 ms │        111.60 / 123.11 ±20.21 / 163.46 ms │   1.09x slower │
│ QQuery 3  │         108.82 / 109.50 ±0.69 / 110.83 ms │         108.99 / 110.03 ±0.60 / 110.82 ms │      no change │
│ QQuery 4  │     1039.25 / 1068.91 ±16.19 / 1085.14 ms │     1039.20 / 1059.08 ±13.72 / 1074.81 ms │      no change │
│ QQuery 5  │         191.03 / 194.29 ±2.37 / 198.33 ms │         193.57 / 197.86 ±3.46 / 202.10 ms │      no change │
│ QQuery 6  │         240.05 / 253.01 ±7.65 / 263.22 ms │         250.53 / 256.69 ±7.61 / 271.10 ms │      no change │
│ QQuery 7  │         335.94 / 339.65 ±3.10 / 343.33 ms │         329.72 / 336.88 ±5.21 / 345.77 ms │      no change │
│ QQuery 8  │         155.31 / 162.03 ±5.23 / 170.94 ms │         154.49 / 163.78 ±4.73 / 167.71 ms │      no change │
│ QQuery 9  │         231.94 / 244.74 ±7.23 / 253.29 ms │        202.44 / 232.73 ±17.86 / 257.46 ms │      no change │
│ QQuery 10 │         168.67 / 173.73 ±4.03 / 179.79 ms │         170.80 / 175.60 ±3.93 / 180.12 ms │      no change │
│ QQuery 11 │         696.40 / 705.47 ±9.70 / 721.59 ms │         687.33 / 698.19 ±6.94 / 708.54 ms │      no change │
│ QQuery 12 │            37.83 / 39.73 ±1.58 / 42.22 ms │            38.72 / 40.03 ±1.68 / 43.35 ms │      no change │
│ QQuery 13 │         554.84 / 564.74 ±8.11 / 572.50 ms │        540.40 / 559.24 ±13.66 / 574.98 ms │      no change │
│ QQuery 14 │         900.79 / 909.28 ±9.25 / 926.53 ms │         910.47 / 918.66 ±5.94 / 927.86 ms │      no change │
│ QQuery 15 │            18.91 / 21.84 ±2.20 / 24.40 ms │            18.89 / 20.28 ±1.49 / 22.93 ms │  +1.08x faster │
│ QQuery 16 │               6.71 / 7.10 ±0.32 / 7.48 ms │               6.82 / 7.23 ±0.28 / 7.62 ms │      no change │
│ QQuery 17 │         204.75 / 206.58 ±2.06 / 210.59 ms │         197.63 / 202.25 ±5.49 / 211.48 ms │      no change │
│ QQuery 18 │        382.30 / 415.20 ±34.38 / 476.68 ms │        424.43 / 481.23 ±32.21 / 514.19 ms │   1.16x slower │
│ QQuery 19 │         141.38 / 145.29 ±2.05 / 147.44 ms │         144.36 / 144.85 ±0.47 / 145.60 ms │      no change │
│ QQuery 20 │            14.98 / 16.95 ±1.18 / 18.19 ms │            14.37 / 16.19 ±1.16 / 17.96 ms │      no change │
│ QQuery 21 │            26.86 / 28.08 ±0.73 / 29.11 ms │            26.02 / 27.64 ±0.94 / 28.91 ms │      no change │
│ QQuery 22 │         497.97 / 505.32 ±4.98 / 511.58 ms │         502.48 / 508.25 ±3.18 / 511.60 ms │      no change │
│ QQuery 23 │      1211.83 / 1218.81 ±4.11 / 1224.08 ms │      1200.17 / 1211.43 ±6.44 / 1218.22 ms │      no change │
│ QQuery 24 │         122.91 / 124.67 ±1.75 / 127.67 ms │         120.61 / 122.13 ±2.17 / 126.36 ms │      no change │
│ QQuery 25 │         293.03 / 302.09 ±7.29 / 311.82 ms │         295.14 / 296.78 ±1.25 / 298.65 ms │      no change │
│ QQuery 26 │         141.27 / 147.20 ±4.80 / 152.46 ms │         138.93 / 144.44 ±3.47 / 147.93 ms │      no change │
│ QQuery 27 │               6.44 / 7.12 ±0.62 / 8.03 ms │              6.59 / 7.60 ±1.69 / 10.96 ms │   1.07x slower │
│ QQuery 28 │         222.05 / 228.22 ±4.98 / 233.80 ms │         224.38 / 229.33 ±3.16 / 234.30 ms │      no change │
│ QQuery 29 │         245.18 / 252.63 ±3.82 / 255.99 ms │         248.74 / 252.39 ±3.05 / 255.90 ms │      no change │
│ QQuery 30 │            55.15 / 58.82 ±2.11 / 61.50 ms │            55.79 / 58.07 ±1.65 / 60.60 ms │      no change │
│ QQuery 31 │         174.66 / 178.17 ±2.47 / 182.08 ms │         173.25 / 175.76 ±1.50 / 177.52 ms │      no change │
│ QQuery 32 │            13.70 / 14.28 ±0.65 / 15.16 ms │            13.30 / 14.02 ±0.65 / 15.18 ms │      no change │
│ QQuery 33 │         128.59 / 132.35 ±2.51 / 136.04 ms │         130.53 / 131.72 ±1.27 / 133.46 ms │      no change │
│ QQuery 34 │               6.73 / 7.24 ±0.51 / 8.07 ms │               6.67 / 7.07 ±0.37 / 7.66 ms │      no change │
│ QQuery 35 │         140.96 / 145.99 ±2.96 / 149.62 ms │         148.59 / 149.43 ±0.89 / 151.03 ms │      no change │
│ QQuery 36 │               6.10 / 6.42 ±0.33 / 6.99 ms │               6.22 / 6.46 ±0.23 / 6.82 ms │      no change │
│ QQuery 37 │               5.06 / 5.23 ±0.12 / 5.36 ms │               5.12 / 5.27 ±0.17 / 5.55 ms │      no change │
│ QQuery 38 │         105.25 / 107.17 ±1.96 / 110.96 ms │         120.95 / 123.98 ±2.58 / 127.69 ms │   1.16x slower │
│ QQuery 39 │         131.88 / 136.51 ±2.69 / 139.14 ms │         135.02 / 138.00 ±2.88 / 143.09 ms │      no change │
│ QQuery 40 │         127.87 / 137.53 ±7.11 / 148.76 ms │         130.45 / 135.30 ±5.64 / 146.34 ms │      no change │
│ QQuery 41 │            13.71 / 14.87 ±1.31 / 17.39 ms │            14.18 / 14.98 ±0.62 / 15.73 ms │      no change │
│ QQuery 42 │         113.89 / 116.31 ±2.30 / 119.77 ms │         108.04 / 112.25 ±2.89 / 116.19 ms │      no change │
│ QQuery 43 │               5.36 / 5.59 ±0.27 / 5.99 ms │               5.46 / 5.61 ±0.16 / 5.81 ms │      no change │
│ QQuery 44 │            11.00 / 12.08 ±0.97 / 13.84 ms │            11.04 / 11.63 ±0.59 / 12.75 ms │      no change │
│ QQuery 45 │            42.40 / 45.35 ±1.71 / 47.53 ms │            42.89 / 46.01 ±2.39 / 49.84 ms │      no change │
│ QQuery 46 │               8.15 / 8.24 ±0.10 / 8.39 ms │               8.20 / 8.63 ±0.25 / 8.94 ms │      no change │
│ QQuery 47 │         770.36 / 773.51 ±3.33 / 778.63 ms │        777.25 / 786.02 ±10.03 / 805.42 ms │      no change │
│ QQuery 48 │         473.85 / 486.47 ±7.43 / 495.18 ms │         479.61 / 486.07 ±4.49 / 493.08 ms │      no change │
│ QQuery 49 │         282.85 / 286.81 ±5.34 / 296.89 ms │         278.60 / 286.90 ±5.09 / 294.61 ms │      no change │
│ QQuery 50 │         531.39 / 536.40 ±3.10 / 541.06 ms │        508.64 / 530.26 ±12.25 / 544.29 ms │      no change │
│ QQuery 51 │         220.57 / 222.52 ±2.08 / 226.25 ms │         214.26 / 217.02 ±2.57 / 221.52 ms │      no change │
│ QQuery 52 │         111.37 / 117.07 ±3.12 / 120.87 ms │         114.24 / 117.26 ±3.72 / 124.46 ms │      no change │
│ QQuery 53 │         139.77 / 146.72 ±5.27 / 153.02 ms │         143.39 / 147.17 ±1.93 / 148.65 ms │      no change │
│ QQuery 54 │         126.09 / 130.28 ±2.13 / 131.81 ms │         128.03 / 131.29 ±2.39 / 134.09 ms │      no change │
│ QQuery 55 │         114.63 / 115.40 ±0.51 / 116.18 ms │         113.90 / 115.50 ±1.03 / 116.68 ms │      no change │
│ QQuery 56 │         131.20 / 134.34 ±1.82 / 135.95 ms │         133.39 / 135.21 ±1.20 / 137.16 ms │      no change │
│ QQuery 57 │         187.17 / 189.15 ±1.33 / 190.51 ms │         187.62 / 188.97 ±0.92 / 190.18 ms │      no change │
│ QQuery 58 │        222.90 / 232.70 ±13.68 / 259.45 ms │         266.45 / 269.46 ±2.02 / 271.86 ms │   1.16x slower │
│ QQuery 59 │         276.92 / 282.59 ±3.74 / 286.75 ms │         278.83 / 281.37 ±1.72 / 283.63 ms │      no change │
│ QQuery 60 │         136.73 / 140.45 ±2.10 / 143.21 ms │         136.29 / 140.31 ±2.93 / 145.05 ms │      no change │
│ QQuery 61 │            11.73 / 12.40 ±0.65 / 13.45 ms │            11.75 / 12.59 ±0.79 / 13.67 ms │      no change │
│ QQuery 62 │       873.89 / 955.85 ±81.38 / 1106.97 ms │       897.57 / 955.56 ±46.81 / 1018.42 ms │      no change │
│ QQuery 63 │         144.49 / 150.98 ±3.84 / 156.47 ms │         140.67 / 146.34 ±3.25 / 150.47 ms │      no change │
│ QQuery 64 │ 28431.72 / 29522.36 ±799.79 / 30673.57 ms │ 27917.80 / 29003.42 ±629.70 / 29529.23 ms │      no change │
│ QQuery 65 │         365.61 / 370.54 ±3.24 / 373.84 ms │         355.82 / 366.45 ±6.16 / 372.19 ms │      no change │
│ QQuery 66 │         190.29 / 197.83 ±4.55 / 201.74 ms │         190.81 / 199.89 ±6.34 / 208.42 ms │      no change │
│ QQuery 67 │        495.23 / 506.92 ±10.04 / 525.59 ms │        471.17 / 497.00 ±14.94 / 515.73 ms │      no change │
│ QQuery 68 │             9.76 / 11.69 ±1.80 / 14.16 ms │             8.23 / 11.23 ±1.69 / 13.03 ms │      no change │
│ QQuery 69 │         171.48 / 178.27 ±4.22 / 182.29 ms │         171.24 / 175.49 ±3.39 / 180.81 ms │      no change │
│ QQuery 70 │         425.58 / 435.99 ±8.02 / 447.51 ms │        415.68 / 432.24 ±11.88 / 450.70 ms │      no change │
│ QQuery 71 │         130.33 / 132.98 ±2.35 / 136.88 ms │         127.03 / 132.91 ±3.40 / 137.09 ms │      no change │
│ QQuery 72 │     1212.21 / 1242.45 ±15.36 / 1254.73 ms │      1180.56 / 1189.01 ±5.13 / 1195.45 ms │      no change │
│ QQuery 73 │               7.14 / 7.82 ±0.67 / 9.02 ms │               7.47 / 7.94 ±0.47 / 8.68 ms │      no change │
│ QQuery 74 │         502.62 / 511.35 ±4.79 / 516.32 ms │         483.17 / 496.74 ±8.78 / 509.24 ms │      no change │
│ QQuery 75 │         280.63 / 286.96 ±4.26 / 292.08 ms │         279.75 / 284.87 ±2.84 / 287.26 ms │      no change │
│ QQuery 76 │         276.35 / 278.08 ±1.16 / 279.11 ms │         271.22 / 277.27 ±3.17 / 280.39 ms │      no change │
│ QQuery 77 │         230.23 / 232.62 ±2.11 / 235.87 ms │         231.14 / 232.05 ±0.91 / 233.71 ms │      no change │
│ QQuery 78 │         290.48 / 301.19 ±6.22 / 309.41 ms │         298.35 / 301.98 ±3.92 / 309.61 ms │      no change │
│ QQuery 79 │         261.13 / 262.32 ±0.99 / 263.54 ms │         260.64 / 263.50 ±1.86 / 266.50 ms │      no change │
│ QQuery 80 │         284.40 / 285.92 ±1.25 / 287.26 ms │         283.91 / 285.71 ±1.19 / 287.26 ms │      no change │
│ QQuery 81 │            31.37 / 33.18 ±2.54 / 38.21 ms │            31.27 / 33.04 ±1.95 / 36.83 ms │      no change │
│ QQuery 82 │            46.07 / 48.60 ±1.42 / 49.71 ms │            48.64 / 49.72 ±1.06 / 51.20 ms │      no change │
│ QQuery 83 │            44.27 / 45.56 ±1.50 / 48.36 ms │            44.65 / 45.67 ±1.11 / 47.71 ms │      no change │
│ QQuery 84 │            65.51 / 66.10 ±0.71 / 67.45 ms │            65.33 / 66.11 ±0.42 / 66.53 ms │      no change │
│ QQuery 85 │         274.71 / 282.53 ±4.55 / 287.69 ms │         282.95 / 295.00 ±6.99 / 304.18 ms │      no change │
│ QQuery 86 │            47.62 / 50.46 ±2.83 / 54.90 ms │            49.51 / 50.97 ±0.87 / 52.07 ms │      no change │
│ QQuery 87 │         105.59 / 107.48 ±1.87 / 110.51 ms │               3.47 / 3.63 ±0.26 / 4.14 ms │ +29.62x faster │
│ QQuery 88 │         127.66 / 132.89 ±3.58 / 138.10 ms │         135.54 / 140.83 ±4.47 / 149.11 ms │   1.06x slower │
│ QQuery 89 │         149.58 / 162.22 ±6.56 / 168.35 ms │         164.05 / 172.94 ±7.63 / 186.40 ms │   1.07x slower │
│ QQuery 90 │            22.94 / 25.05 ±1.42 / 26.50 ms │            26.99 / 28.59 ±0.95 / 29.67 ms │   1.14x slower │
│ QQuery 91 │          98.77 / 105.49 ±5.88 / 113.94 ms │         103.94 / 110.26 ±3.23 / 113.12 ms │      no change │
│ QQuery 92 │            53.08 / 53.86 ±0.67 / 54.77 ms │            54.87 / 57.00 ±1.38 / 58.80 ms │   1.06x slower │
│ QQuery 93 │         186.64 / 188.65 ±1.11 / 190.01 ms │         196.58 / 199.04 ±1.44 / 200.54 ms │   1.06x slower │
│ QQuery 94 │            67.22 / 68.67 ±1.41 / 71.10 ms │            68.87 / 70.88 ±1.25 / 72.53 ms │      no change │
│ QQuery 95 │         144.69 / 147.21 ±1.56 / 149.55 ms │         143.71 / 148.91 ±3.15 / 152.91 ms │      no change │
│ QQuery 96 │            78.21 / 85.24 ±4.06 / 88.74 ms │            80.83 / 87.93 ±3.77 / 91.35 ms │      no change │
│ QQuery 97 │         149.40 / 154.35 ±2.83 / 157.03 ms │         150.98 / 154.44 ±3.29 / 160.14 ms │      no change │
│ QQuery 98 │         125.07 / 126.22 ±1.46 / 129.06 ms │         122.50 / 125.05 ±2.23 / 128.53 ms │      no change │
│ QQuery 99 │  10658.19 / 10710.06 ±27.14 / 10734.49 ms │  10705.42 / 10732.23 ±23.33 / 10770.96 ms │      no change │
└───────────┴───────────────────────────────────────────┴───────────────────────────────────────────┴────────────────┘
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━┓
┃ Benchmark Summary                             ┃            ┃
┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━┩
│ Total Time (HEAD)                             │ 61916.72ms │
│ Total Time (use-ndv-for-semi-and-anti-join)   │ 61371.94ms │
│ Average Time (HEAD)                           │   625.42ms │
│ Average Time (use-ndv-for-semi-and-anti-join) │   619.92ms │
│ Queries Faster                                │          2 │
│ Queries Slower                                │         10 │
│ Queries with No Change                        │         87 │
│ Queries with Failure                          │          0 │
└───────────────────────────────────────────────┴────────────┘

Resource Usage

tpcds — base (merge-base)

Metric	Value
Wall time	310.1s
Peak memory	37.6 GiB
Avg memory	26.1 GiB
CPU user	606.8s
CPU sys	27.3s
Peak spill	0 B

tpcds — branch

Metric	Value
Wall time	310.1s
Peak memory	35.4 GiB
Avg memory	26.0 GiB
CPU user	608.1s
CPU sys	26.5s
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 use-ndv-for-semi-and-anti-join
--------------------
Benchmark tpcds_sf1.json
--------------------
┏━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━┓
┃ Query     ┃                                     HEAD ┃           use-ndv-for-semi-and-anti-join ┃         Change ┃
┡━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━┩
│ QQuery 1  │              6.50 / 6.93 ±0.72 / 8.37 ms │              6.50 / 6.88 ±0.70 / 8.29 ms │      no change │
│ QQuery 2  │        147.80 / 148.79 ±0.72 / 149.98 ms │        145.90 / 147.29 ±1.00 / 148.59 ms │      no change │
│ QQuery 3  │        114.64 / 115.80 ±1.18 / 118.00 ms │        113.74 / 114.29 ±0.52 / 115.13 ms │      no change │
│ QQuery 4  │    1309.33 / 1330.70 ±12.99 / 1347.24 ms │    1296.30 / 1322.23 ±17.58 / 1345.81 ms │      no change │
│ QQuery 5  │        173.18 / 173.93 ±0.41 / 174.31 ms │        173.98 / 175.64 ±1.46 / 177.94 ms │      no change │
│ QQuery 6  │       827.65 / 851.71 ±21.91 / 885.61 ms │       845.56 / 866.91 ±17.66 / 887.56 ms │      no change │
│ QQuery 7  │        340.37 / 344.31 ±2.22 / 346.95 ms │        339.07 / 343.20 ±3.84 / 350.04 ms │      no change │
│ QQuery 8  │        116.01 / 117.89 ±1.15 / 119.26 ms │        114.14 / 116.70 ±1.49 / 118.35 ms │      no change │
│ QQuery 9  │        101.97 / 107.00 ±4.74 / 114.97 ms │        102.87 / 105.69 ±1.44 / 106.82 ms │      no change │
│ QQuery 10 │        105.45 / 106.43 ±0.72 / 107.24 ms │        104.56 / 105.86 ±1.04 / 107.15 ms │      no change │
│ QQuery 11 │        930.44 / 943.03 ±7.99 / 954.30 ms │       929.20 / 943.86 ±11.12 / 955.59 ms │      no change │
│ QQuery 12 │           44.83 / 45.61 ±0.58 / 46.26 ms │           44.43 / 46.20 ±1.37 / 48.20 ms │      no change │
│ QQuery 13 │        396.18 / 400.46 ±2.60 / 403.51 ms │        402.21 / 405.79 ±2.24 / 409.11 ms │      no change │
│ QQuery 14 │     1008.21 / 1016.16 ±7.94 / 1030.03 ms │     1006.52 / 1009.43 ±3.33 / 1015.62 ms │      no change │
│ QQuery 15 │           15.52 / 16.53 ±1.23 / 18.94 ms │           15.22 / 15.93 ±0.69 / 17.20 ms │      no change │
│ QQuery 16 │              7.33 / 8.22 ±0.94 / 9.71 ms │              7.25 / 8.12 ±0.89 / 9.35 ms │      no change │
│ QQuery 17 │        230.28 / 230.39 ±0.10 / 230.52 ms │        228.54 / 231.89 ±1.75 / 233.43 ms │      no change │
│ QQuery 18 │        127.04 / 128.06 ±1.09 / 130.04 ms │        127.05 / 128.72 ±1.31 / 130.45 ms │      no change │
│ QQuery 19 │        154.48 / 155.70 ±0.99 / 156.82 ms │        155.05 / 157.13 ±1.26 / 158.91 ms │      no change │
│ QQuery 20 │           13.23 / 14.04 ±0.70 / 15.05 ms │           13.51 / 14.22 ±0.56 / 15.06 ms │      no change │
│ QQuery 21 │           19.31 / 19.71 ±0.25 / 20.01 ms │           18.92 / 19.80 ±0.59 / 20.62 ms │      no change │
│ QQuery 22 │        484.33 / 487.03 ±1.86 / 489.23 ms │        483.55 / 486.45 ±2.75 / 491.19 ms │      no change │
│ QQuery 23 │        871.93 / 881.92 ±7.27 / 892.37 ms │        861.04 / 872.89 ±8.74 / 885.04 ms │      no change │
│ QQuery 24 │        381.29 / 384.84 ±2.51 / 388.11 ms │        381.28 / 382.82 ±1.90 / 386.40 ms │      no change │
│ QQuery 25 │        341.44 / 342.93 ±1.11 / 344.51 ms │        338.52 / 341.34 ±2.16 / 344.45 ms │      no change │
│ QQuery 26 │           81.55 / 82.83 ±1.45 / 84.85 ms │           80.75 / 82.27 ±0.86 / 83.30 ms │      no change │
│ QQuery 27 │              6.75 / 7.44 ±0.61 / 8.25 ms │              6.86 / 7.30 ±0.33 / 7.83 ms │      no change │
│ QQuery 28 │        149.28 / 150.16 ±0.89 / 151.54 ms │        149.81 / 151.51 ±1.31 / 153.86 ms │      no change │
│ QQuery 29 │        280.30 / 281.92 ±1.82 / 285.20 ms │        280.01 / 282.04 ±1.83 / 285.22 ms │      no change │
│ QQuery 30 │           43.40 / 45.76 ±1.35 / 47.56 ms │           43.04 / 46.04 ±1.61 / 47.59 ms │      no change │
│ QQuery 31 │        169.93 / 171.64 ±1.01 / 172.55 ms │        171.87 / 173.18 ±1.53 / 176.15 ms │      no change │
│ QQuery 32 │           13.85 / 14.77 ±1.08 / 16.77 ms │           13.48 / 13.83 ±0.20 / 14.08 ms │  +1.07x faster │
│ QQuery 33 │        139.11 / 142.24 ±1.95 / 145.05 ms │        138.87 / 141.48 ±1.80 / 143.82 ms │      no change │
│ QQuery 34 │              7.03 / 7.43 ±0.45 / 8.26 ms │              6.82 / 7.10 ±0.24 / 7.51 ms │      no change │
│ QQuery 35 │        108.43 / 109.93 ±1.27 / 111.79 ms │        103.08 / 108.13 ±2.62 / 110.55 ms │      no change │
│ QQuery 36 │              6.79 / 7.32 ±0.80 / 8.91 ms │              6.39 / 6.92 ±0.50 / 7.86 ms │  +1.06x faster │
│ QQuery 37 │              8.34 / 8.78 ±0.39 / 9.31 ms │             8.70 / 9.25 ±0.64 / 10.49 ms │   1.05x slower │
│ QQuery 38 │           84.39 / 89.84 ±4.76 / 96.78 ms │           93.69 / 97.43 ±2.14 / 99.90 ms │   1.08x slower │
│ QQuery 39 │        125.05 / 129.19 ±2.53 / 132.53 ms │        123.39 / 127.13 ±2.09 / 129.45 ms │      no change │
│ QQuery 40 │        114.18 / 117.69 ±5.40 / 128.36 ms │        108.82 / 114.30 ±6.68 / 126.21 ms │      no change │
│ QQuery 41 │           14.31 / 14.97 ±0.33 / 15.20 ms │           14.01 / 15.00 ±0.71 / 16.14 ms │      no change │
│ QQuery 42 │        108.44 / 109.73 ±1.07 / 110.97 ms │        107.07 / 109.18 ±1.48 / 111.11 ms │      no change │
│ QQuery 43 │              5.57 / 5.73 ±0.18 / 6.08 ms │              5.60 / 5.79 ±0.16 / 6.03 ms │      no change │
│ QQuery 44 │           11.80 / 11.95 ±0.19 / 12.32 ms │           11.56 / 11.98 ±0.37 / 12.44 ms │      no change │
│ QQuery 45 │           51.75 / 52.29 ±0.81 / 53.90 ms │           51.11 / 51.42 ±0.38 / 52.17 ms │      no change │
│ QQuery 46 │              8.29 / 8.53 ±0.25 / 8.96 ms │              8.16 / 8.48 ±0.21 / 8.80 ms │      no change │
│ QQuery 47 │        720.53 / 727.27 ±3.99 / 731.64 ms │        707.26 / 717.71 ±5.98 / 724.63 ms │      no change │
│ QQuery 48 │        288.90 / 296.87 ±4.51 / 301.23 ms │        285.94 / 291.99 ±4.90 / 296.83 ms │      no change │
│ QQuery 49 │        254.16 / 258.22 ±2.87 / 262.48 ms │        253.03 / 255.36 ±1.78 / 257.94 ms │      no change │
│ QQuery 50 │        226.95 / 230.73 ±2.92 / 234.47 ms │        213.17 / 223.62 ±7.11 / 233.07 ms │      no change │
│ QQuery 51 │        179.47 / 184.37 ±3.14 / 189.20 ms │        180.79 / 182.36 ±1.65 / 185.27 ms │      no change │
│ QQuery 52 │        107.85 / 108.61 ±0.85 / 110.25 ms │        107.59 / 108.58 ±0.67 / 109.62 ms │      no change │
│ QQuery 53 │        103.08 / 104.97 ±1.50 / 106.65 ms │        102.44 / 104.44 ±1.62 / 106.40 ms │      no change │
│ QQuery 54 │        146.42 / 147.50 ±1.03 / 149.35 ms │        143.55 / 146.46 ±1.59 / 148.37 ms │      no change │
│ QQuery 55 │        107.25 / 107.92 ±0.88 / 109.65 ms │        106.99 / 108.64 ±1.33 / 110.32 ms │      no change │
│ QQuery 56 │        140.60 / 142.68 ±1.53 / 144.36 ms │        141.34 / 142.84 ±1.25 / 144.62 ms │      no change │
│ QQuery 57 │        170.93 / 174.74 ±2.89 / 177.28 ms │        171.09 / 173.89 ±1.54 / 175.14 ms │      no change │
│ QQuery 58 │        274.48 / 275.85 ±0.73 / 276.57 ms │        318.23 / 320.04 ±1.39 / 321.84 ms │   1.16x slower │
│ QQuery 59 │        199.15 / 199.93 ±0.91 / 201.57 ms │        196.75 / 198.63 ±1.00 / 199.40 ms │      no change │
│ QQuery 60 │        142.53 / 145.07 ±2.13 / 148.08 ms │        142.20 / 143.94 ±1.67 / 146.64 ms │      no change │
│ QQuery 61 │           13.06 / 13.22 ±0.09 / 13.34 ms │           12.70 / 12.91 ±0.28 / 13.45 ms │      no change │
│ QQuery 62 │       883.24 / 920.83 ±24.13 / 948.13 ms │       877.61 / 933.48 ±40.77 / 989.54 ms │      no change │
│ QQuery 63 │        104.85 / 108.45 ±2.56 / 111.33 ms │        105.01 / 107.69 ±1.98 / 110.08 ms │      no change │
│ QQuery 64 │        685.34 / 690.86 ±4.84 / 697.70 ms │        683.99 / 689.11 ±4.08 / 696.22 ms │      no change │
│ QQuery 65 │        250.62 / 254.71 ±3.26 / 260.42 ms │        248.16 / 250.93 ±2.43 / 254.53 ms │      no change │
│ QQuery 66 │        250.46 / 255.59 ±3.16 / 259.69 ms │       237.07 / 250.19 ±11.39 / 265.08 ms │      no change │
│ QQuery 67 │        310.59 / 316.52 ±4.86 / 323.93 ms │       304.83 / 315.90 ±11.03 / 336.78 ms │      no change │
│ QQuery 68 │             8.36 / 9.05 ±0.74 / 10.48 ms │            8.62 / 10.86 ±1.52 / 12.58 ms │   1.20x slower │
│ QQuery 69 │        102.21 / 103.49 ±1.25 / 105.04 ms │        101.55 / 103.58 ±1.29 / 105.47 ms │      no change │
│ QQuery 70 │       334.60 / 347.02 ±12.14 / 370.30 ms │       331.40 / 350.51 ±13.42 / 373.15 ms │      no change │
│ QQuery 71 │        135.31 / 137.17 ±1.65 / 139.25 ms │        135.48 / 137.10 ±1.68 / 140.30 ms │      no change │
│ QQuery 72 │        615.01 / 619.88 ±5.32 / 629.37 ms │       614.82 / 627.97 ±10.29 / 644.25 ms │      no change │
│ QQuery 73 │              6.70 / 7.75 ±0.85 / 8.90 ms │             6.55 / 8.02 ±1.22 / 10.23 ms │      no change │
│ QQuery 74 │        584.37 / 593.13 ±6.95 / 601.45 ms │        563.40 / 573.94 ±6.44 / 581.98 ms │      no change │
│ QQuery 75 │        273.01 / 277.45 ±2.87 / 281.43 ms │        275.58 / 278.67 ±2.50 / 281.48 ms │      no change │
│ QQuery 76 │        131.80 / 133.42 ±0.97 / 134.50 ms │        131.12 / 133.31 ±2.13 / 136.05 ms │      no change │
│ QQuery 77 │        189.32 / 190.42 ±1.35 / 193.06 ms │        187.48 / 189.90 ±2.36 / 194.21 ms │      no change │
│ QQuery 78 │        341.40 / 344.94 ±3.24 / 349.15 ms │        337.74 / 342.76 ±3.22 / 346.79 ms │      no change │
│ QQuery 79 │        231.49 / 235.69 ±3.39 / 241.77 ms │        230.20 / 234.24 ±3.23 / 238.14 ms │      no change │
│ QQuery 80 │        323.06 / 327.51 ±2.75 / 330.97 ms │        323.85 / 326.29 ±3.05 / 332.13 ms │      no change │
│ QQuery 81 │           25.79 / 27.15 ±1.19 / 28.89 ms │           26.14 / 26.82 ±0.68 / 28.01 ms │      no change │
│ QQuery 82 │           39.71 / 41.33 ±1.20 / 43.08 ms │           41.14 / 41.49 ±0.35 / 42.11 ms │      no change │
│ QQuery 83 │           38.53 / 39.12 ±0.46 / 39.94 ms │           37.89 / 39.54 ±1.65 / 42.60 ms │      no change │
│ QQuery 84 │           48.13 / 49.02 ±1.05 / 50.91 ms │           48.30 / 49.86 ±1.34 / 52.34 ms │      no change │
│ QQuery 85 │        147.75 / 149.25 ±1.12 / 150.75 ms │        147.33 / 149.64 ±2.58 / 154.62 ms │      no change │
│ QQuery 86 │           38.36 / 39.64 ±0.80 / 40.65 ms │           38.61 / 39.14 ±0.34 / 39.51 ms │      no change │
│ QQuery 87 │           86.85 / 89.70 ±3.00 / 95.07 ms │              3.36 / 3.48 ±0.19 / 3.87 ms │ +25.77x faster │
│ QQuery 88 │         99.89 / 101.01 ±1.40 / 103.55 ms │           98.34 / 99.05 ±0.42 / 99.64 ms │      no change │
│ QQuery 89 │        120.01 / 121.03 ±0.84 / 122.41 ms │        117.76 / 119.00 ±1.13 / 121.01 ms │      no change │
│ QQuery 90 │           23.86 / 24.34 ±0.50 / 25.29 ms │           22.75 / 23.44 ±0.37 / 23.83 ms │      no change │
│ QQuery 91 │           62.53 / 65.14 ±2.30 / 69.26 ms │           62.19 / 64.26 ±2.23 / 67.90 ms │      no change │
│ QQuery 92 │           58.08 / 58.90 ±0.57 / 59.67 ms │           57.37 / 58.19 ±0.70 / 59.24 ms │      no change │
│ QQuery 93 │        185.33 / 188.45 ±1.76 / 189.98 ms │        185.93 / 188.09 ±1.43 / 189.80 ms │      no change │
│ QQuery 94 │           62.90 / 63.34 ±0.37 / 63.86 ms │           62.30 / 63.62 ±0.99 / 64.90 ms │      no change │
│ QQuery 95 │        131.50 / 133.31 ±1.40 / 134.53 ms │        130.26 / 131.68 ±1.20 / 133.51 ms │      no change │
│ QQuery 96 │           71.67 / 74.31 ±1.40 / 75.74 ms │           72.93 / 73.58 ±0.80 / 74.96 ms │      no change │
│ QQuery 97 │        126.24 / 128.50 ±2.02 / 131.78 ms │        125.38 / 128.41 ±1.96 / 130.93 ms │      no change │
│ QQuery 98 │        153.05 / 154.07 ±0.58 / 154.65 ms │        150.15 / 153.68 ±2.56 / 157.40 ms │      no change │
│ QQuery 99 │ 10742.43 / 10782.57 ±34.38 / 10834.75 ms │ 10731.90 / 10754.02 ±16.29 / 10777.09 ms │      no change │
└───────────┴──────────────────────────────────────────┴──────────────────────────────────────────┴────────────────┘
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━┓
┃ Benchmark Summary                             ┃            ┃
┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━┩
│ Total Time (HEAD)                             │ 31270.23ms │
│ Total Time (use-ndv-for-semi-and-anti-join)   │ 31141.86ms │
│ Average Time (HEAD)                           │   315.86ms │
│ Average Time (use-ndv-for-semi-and-anti-join) │   314.56ms │
│ Queries Faster                                │          3 │
│ Queries Slower                                │          4 │
│ Queries with No Change                        │         92 │
│ Queries with Failure                          │          0 │
└───────────────────────────────────────────────┴────────────┘

Resource Usage

tpcds — base (merge-base)

Metric	Value
Wall time	160.0s
Peak memory	6.7 GiB
Avg memory	5.6 GiB
CPU user	259.9s
CPU sys	16.7s
Peak spill	0 B

tpcds — branch

Metric	Value
Wall time	160.0s
Peak memory	6.6 GiB
Avg memory	5.5 GiB
CPU user	256.7s
CPU sys	16.8s
Peak spill	0 B

---

File an issue against this benchmark runner

[xudong963]

│ QQuery 87 │ 85.75 / 88.07 ±2.62 / 93.17 ms │ 3.60 / 3.80 ±0.19 / 4.09 ms │ +23.19x faster │

Interesting!

The magic of stats!

[xudong963]

select count(*) 
from ((select distinct c_last_name, c_first_name, d_date
       from store_sales, date_dim, customer
       where store_sales.ss_sold_date_sk = date_dim.d_date_sk
         and store_sales.ss_customer_sk = customer.c_customer_sk
         and d_month_seq between 1189 and 1189+11)
       except
      (select distinct c_last_name, c_first_name, d_date
       from catalog_sales, date_dim, customer
       where catalog_sales.cs_sold_date_sk = date_dim.d_date_sk
         and catalog_sales.cs_bill_customer_sk = customer.c_customer_sk
         and d_month_seq between 1189 and 1189+11)
       except
      (select distinct c_last_name, c_first_name, d_date
       from web_sales, date_dim, customer
       where web_sales.ws_sold_date_sk = date_dim.d_date_sk
         and web_sales.ws_bill_customer_sk = customer.c_customer_sk
         and d_month_seq between 1189 and 1189+11)
) cool_cust
;

I didn't find semi-anti joins for the query. I guess it might be related to the count(*) gets a stat directly

[adriangbot]

Benchmark for this request hit the 7200s job deadline before finishing.

Benchmarks requested: tpch

Kubernetes message

Job was active longer than specified deadline

---

File an issue against this benchmark runner

[adriangbot]

Benchmark for this request hit the 7200s job deadline before finishing.

Benchmarks requested: tpch

Kubernetes message

Job was active longer than specified deadline

---

File an issue against this benchmark runner

[adriangbot]

Benchmark for this request hit the 7200s job deadline before finishing.

Benchmarks requested: tpch

Kubernetes message

Job was active longer than specified deadline

---

File an issue against this benchmark runner