Skip to content

Commit 5c24327

Browse files
github-actions[bot]noahgift
github-actions[bot]
and
noahgift
authored
bench: criterion results for 1005a55 (#14)
* bench: criterion results for 1005a55 Auto-generated by .github/workflows/bench.yml on intel-clean-room. Triggered by workflow_dispatch. See bench-results/latest/SUMMARY.md for the rows/sec table. HTML report attached as workflow artifact. * ci: empty commit to trigger gate check --------- Co-authored-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com> Co-authored-by: Noah Gift <noah.gift@gmail.com>
1 parent 1005a55 commit 5c24327

5 files changed

Lines changed: 26 additions & 60 deletions

File tree

‎bench-results/latest/SUMMARY.md‎

Lines changed: 16 additions & 49 deletions
Original file line numberDiff line numberDiff line change
@@ -1,52 +1,19 @@
1-
# Criterion bench results — 719f76c
2-
3-
```
4-
host=noah-Lambda-Vector
5-
os=Linux 6.8.0-90-generic x86_64
6-
cpu=AMD Ryzen Threadripper 7960X 24-Cores
7-
cores=48
8-
rustc=rustc 1.95.0 (59807616e 2026-04-14)
9-
cargo=cargo 1.95.0 (f2d3ce0bd 2026-03-21)
10-
git_sha=719f76c
11-
git_ref=refs/heads/feat/bench-workflow
12-
ts=2026-04-28T10:58:41Z
13-
trigger=local-seed-run
14-
```
1+
# Criterion bench results — 1005a55
2+
3+
host=mac-server
4+
os=Linux 6.8.0-107-generic x86_64
5+
cpu=Intel(R) Xeon(R) W-3245 CPU @ 3.20GHz
6+
cores=32
7+
rustc=rustc 1.95.0 (59807616e 2026-04-14)
8+
cargo=cargo 1.95.0 (f2d3ce0bd 2026-03-21)
9+
git_sha=1005a55
10+
git_ref=refs/heads/main
11+
ts=2026-04-28T14:36:22Z
1512

1613
## Throughput (rows/sec)
1714

18-
| Size | Mean (ns) | Std Dev (ns) | Throughput |
19-
|------|-----------|--------------|------------|
20-
| 1000 | 111,057 | 4,348 | 9.00 Melem/s |
21-
| 10000 | 1,025,980 | 27,510 | 9.75 Melem/s |
22-
| 100000 | 10,375,079 | 286,954 | 9.64 Melem/s |
23-
24-
## How to read these results
25-
26-
Each row reports the mean wall-clock time per `etl_core::run` call
27-
across the full input (1k / 10k / 100k synthetic CSV rows). The
28-
throughput column is `size / mean` — what matters is that it stays
29-
roughly constant across sizes (it does, ~9.6 Melem/s here) which
30-
confirms the pipeline is **linear** in input size.
31-
32-
Per-call overhead (CSV reader init, allocation) is amortized away by
33-
10k rows; the 1k size catches startup-cost regressions.
34-
35-
## Reproducing locally
36-
37-
```bash
38-
make bench # full criterion suite (~30s)
39-
make bench-smoke # CI smoke mode (compile + run-once)
40-
```
41-
42-
## Automated re-runs
43-
44-
[`.github/workflows/bench.yml`](../.github/workflows/bench.yml) runs the
45-
full suite on a self-hosted `intel` runner (paiml org runner pool):
46-
47-
- on `workflow_dispatch` (manual)
48-
- weekly via cron (Sundays 06:00 UTC)
49-
- on push to `main` when `etl-core/`, `etl-bench/`, or `Cargo.{toml,lock}` change
50-
51-
Results are committed back to this directory by the workflow.
52-
The HTML report is also available as a workflow artifact.
15+
| Size | Mean | Std Dev | Throughput |
16+
|------|------|---------|------------|
17+
| 1000 | 163011 ns | 1372 ns | 6134555 rows/sec |
18+
| 10000 | 1515682 ns | 40834 ns | 6597688 rows/sec |
19+
| 100000 | 15159020 ns | 172458 ns | 6596732 rows/sec |

‎bench-results/latest/etl_throughput/1000/estimates.json‎

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -1 +1 @@
1-
{"mean":{"confidence_interval":{"confidence_level":0.95,"lower_bound":110232.45128950721,"upper_bound":111957.02887769202},"point_estimate":111057.84693852508,"standard_error":434.0713279255122},"median":{"confidence_interval":{"confidence_level":0.95,"lower_bound":110151.55317460318,"upper_bound":111315.73784722222},"point_estimate":110572.46313164172,"standard_error":312.7989708707826},"median_abs_dev":{"confidence_interval":{"confidence_level":0.95,"lower_bound":2373.0077883201043,"upper_bound":3693.427244259129},"point_estimate":3070.3579430337045,"standard_error":342.9713276081519},"slope":{"confidence_interval":{"confidence_level":0.95,"lower_bound":110877.85361878379,"upper_bound":112306.29055909732},"point_estimate":111586.76540663022,"standard_error":364.78391695034054},"std_dev":{"confidence_interval":{"confidence_level":0.95,"lower_bound":3216.5390427476436,"upper_bound":5488.964147083875},"point_estimate":4348.216659920425,"standard_error":590.1900150236157}}
1+
{"mean":{"confidence_interval":{"confidence_level":0.95,"lower_bound":162734.15718955308,"upper_bound":163268.05833638087},"point_estimate":163011.01837201524,"standard_error":135.07619077013496},"median":{"confidence_interval":{"confidence_level":0.95,"lower_bound":163528.1484425349,"upper_bound":163891.77265745008},"point_estimate":163800.88844155843,"standard_error":125.05257291001152},"median_abs_dev":{"confidence_interval":{"confidence_level":0.95,"lower_bound":422.2130740831771,"upper_bound":1146.3920982958339},"point_estimate":621.2349010442898,"standard_error":199.57097644901353},"slope":{"confidence_interval":{"confidence_level":0.95,"lower_bound":162102.06822504057,"upper_bound":162728.71164201785},"point_estimate":162413.5287774705,"standard_error":159.93522505770014},"std_dev":{"confidence_interval":{"confidence_level":0.95,"lower_bound":1203.3245145637334,"upper_bound":1506.2675567019428},"point_estimate":1371.7707594278077,"standard_error":77.25100077809408}}

‎bench-results/latest/etl_throughput/10000/estimates.json‎

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -1 +1 @@
1-
{"mean":{"confidence_interval":{"confidence_level":0.95,"lower_bound":1020576.8120204487,"upper_bound":1031349.7867290315},"point_estimate":1025980.9293549486,"standard_error":2732.904864634797},"median":{"confidence_interval":{"confidence_level":0.95,"lower_bound":1021893.25,"upper_bound":1034256.8192528735},"point_estimate":1027993.1523148149,"standard_error":3149.744984407961},"median_abs_dev":{"confidence_interval":{"confidence_level":0.95,"lower_bound":15625.22347223641,"upper_bound":30555.90771397906},"point_estimate":20240.219507330985,"standard_error":3833.2738773654787},"slope":{"confidence_interval":{"confidence_level":0.95,"lower_bound":1027633.1976448015,"upper_bound":1040304.3068307431},"point_estimate":1034135.4093719522,"standard_error":3231.5640124900474},"std_dev":{"confidence_interval":{"confidence_level":0.95,"lower_bound":23261.676879059716,"upper_bound":31457.12901860668},"point_estimate":27510.88292907078,"standard_error":2076.422750542962}}
1+
{"mean":{"confidence_interval":{"confidence_level":0.95,"lower_bound":1509003.6668665388,"upper_bound":1524676.1390615029},"point_estimate":1515682.3784219637,"standard_error":4063.663893094052},"median":{"confidence_interval":{"confidence_level":0.95,"lower_bound":1509886.5107334526,"upper_bound":1511437.2857142857},"point_estimate":1510503.7455197133,"standard_error":343.5089445248264},"median_abs_dev":{"confidence_interval":{"confidence_level":0.95,"lower_bound":3134.2140421826216,"upper_bound":9425.89163881089},"point_estimate":4816.877978329449,"standard_error":1626.4161496983122},"slope":{"confidence_interval":{"confidence_level":0.95,"lower_bound":1499207.0937789169,"upper_bound":1513115.5506409206},"point_estimate":1505623.1271760012,"standard_error":3570.4356238910345},"std_dev":{"confidence_interval":{"confidence_level":0.95,"lower_bound":18253.716597088263,"upper_bound":63693.17423928173},"point_estimate":40834.40033150195,"standard_error":14249.539047487151}}

‎bench-results/latest/etl_throughput/100000/estimates.json‎

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -1 +1 @@
1-
{"mean":{"confidence_interval":{"confidence_level":0.95,"lower_bound":10318149.482399998,"upper_bound":10430364.186},"point_estimate":10375079.899999999,"standard_error":28544.392728229002},"median":{"confidence_interval":{"confidence_level":0.95,"lower_bound":10330161.6,"upper_bound":10417047.2},"point_estimate":10376796.2,"standard_error":23856.858123839913},"median_abs_dev":{"confidence_interval":{"confidence_level":0.95,"lower_bound":170196.25005841366,"upper_bound":316630.8703386792},"point_estimate":228123.35841000162,"standard_error":37726.92208440594},"slope":null,"std_dev":{"confidence_interval":{"confidence_level":0.95,"lower_bound":239036.35582141922,"upper_bound":331659.36838515924},"point_estimate":286954.78517900995,"standard_error":23666.22269342265}}
1+
{"mean":{"confidence_interval":{"confidence_level":0.95,"lower_bound":15125904.8325,"upper_bound":15192579.6255},"point_estimate":15159020.205,"standard_error":17082.613466955576},"median":{"confidence_interval":{"confidence_level":0.95,"lower_bound":15094417.0,"upper_bound":15196464.0},"point_estimate":15136136.25,"standard_error":29267.63068704696},"median_abs_dev":{"confidence_interval":{"confidence_level":0.95,"lower_bound":148040.57257175446,"upper_bound":227787.79970470793},"point_estimate":191450.72915107012,"standard_error":21092.02780823823},"slope":null,"std_dev":{"confidence_interval":{"confidence_level":0.95,"lower_bound":151041.07097649088,"upper_bound":191573.44563334252},"point_estimate":172458.25136017363,"standard_error":10372.631969824159}}

‎bench-results/latest/meta.txt‎

Lines changed: 7 additions & 8 deletions
Original file line numberDiff line numberDiff line change
@@ -1,10 +1,9 @@
1-
host=noah-Lambda-Vector
2-
os=Linux 6.8.0-90-generic x86_64
3-
cpu=AMD Ryzen Threadripper 7960X 24-Cores
4-
cores=48
1+
host=mac-server
2+
os=Linux 6.8.0-107-generic x86_64
3+
cpu=Intel(R) Xeon(R) W-3245 CPU @ 3.20GHz
4+
cores=32
55
rustc=rustc 1.95.0 (59807616e 2026-04-14)
66
cargo=cargo 1.95.0 (f2d3ce0bd 2026-03-21)
7-
git_sha=719f76c
8-
git_ref=refs/heads/feat/bench-workflow
9-
ts=2026-04-28T10:58:41Z
10-
trigger=local-seed-run
7+
git_sha=1005a55
8+
git_ref=refs/heads/main
9+
ts=2026-04-28T14:36:22Z

0 commit comments

Comments
 (0)