Skip to content
KernelIndex
Search⌘K
Sign in

Causal depthwise conv1d

36 eligible runs
conv

Causal depthwise 1D convolution, a core component of Mamba/Mamba-2 architectures: each channel is convolved independently with causal left zero-padding, so out[b, d, t] = bias[d] + sum_k weight[d, k] * x[b, d, t-w+1+k]. All tensors are fp32.

Fastest reported
10.0µsgeomean
submission 555004brandonin · License unknown · python

Reported evidence · last observed 2026-03-14. Reported by source; not independently reproduced. License unknown.

No package · vendor the source →View source →Run detail →

Current records

Not measured on H100 for this workload. Challenges →

Source-native comparison · GPU NVIDIA B200 · Workload suite of 3 cases · geomean score · Protocol reference-kernels eval.py · geomean · 36 results · last observed 2026-03-15Record history →
#
Implementation
Latency
vs #1
Trust
Observed
1
submission 555004brandonin
10.0µs
1.00×
Reported · license unknown · source
2026-03-14stale

Measured exactly what you asked. Reported by source; not independently reproduced.

source mirroredlicense unknownno install recipeView source →Run detail →
2
submission 555480Ayush10
12.6µs
1.25×
Reported · license unknown · source
2026-03-15stale

Measured exactly what you asked. Reported by source; not independently reproduced.

source mirroredlicense unknownno install recipeView source →Run detail →
3
submission 553892Mistobaan
13.6µs
1.35×
Reported · license unknown · source
2026-03-14stale

Measured exactly what you asked. Reported by source; not independently reproduced.

source mirroredlicense unknownno install recipeView source →Run detail →
4
submission 554279Ayush10
13.9µs
1.38×
Reported · license unknown · source
2026-03-14stale

Measured exactly what you asked. Reported by source; not independently reproduced.

source mirroredlicense unknownno install recipeView source →Run detail →
5
submission 554383CodingMaster
14.0µs
1.39×
Reported · license unknown · source
2026-03-14stale

Measured exactly what you asked. Reported by source; not independently reproduced.

source mirroredlicense unknownno install recipeView source →Run detail →
6
submission 554374Narain
15.0µs
1.49×
Reported · license unknown · source
2026-03-14stale

Measured exactly what you asked. Reported by source; not independently reproduced.

source mirroredlicense unknownno install recipeView source →Run detail →
7
submission 554984kitrak_rev.
17.3µs
1.72×
Reported · license unknown · source
2026-03-14stale

Measured exactly what you asked. Reported by source; not independently reproduced.

source mirroredlicense unknownno install recipeView source →Run detail →
8
submission 555604the.defenestrator
19.9µs
1.98×
Reported · license unknown · source
2026-03-15stale

Measured exactly what you asked. Reported by source; not independently reproduced.

source mirroredlicense unknownno install recipeView source →Run detail →
9
submission 555492sissi1165
22.1µs
2.20×
Reported · license unknown · source
2026-03-15stale

Measured exactly what you asked. Reported by source; not independently reproduced.

source mirroredlicense unknownno install recipeView source →Run detail →
10
submission 555503alazarr.m
23.9µs
2.38×
Reported · license unknown · source
2026-03-15stale

Measured exactly what you asked. Reported by source; not independently reproduced.

source mirroredlicense unknownno install recipeView source →Run detail →
11
submission 555590fluudgate
25.1µs
2.50×
Reported · license unknown · source
2026-03-15stale

Measured exactly what you asked. Reported by source; not independently reproduced.

source mirroredlicense unknownno install recipeView source →Run detail →
12
submission 555174plops655
25.4µs
2.53×
Reported · license unknown · source
2026-03-15stale

Measured exactly what you asked. Reported by source; not independently reproduced.

source mirroredlicense unknownno install recipeView source →Run detail →
13
submission 555131dpang
26.0µs
2.58×
Reported · license unknown · source
2026-03-15stale

Measured exactly what you asked. Reported by source; not independently reproduced.

source mirroredlicense unknownno install recipeView source →Run detail →
14
submission 552586yhinai
26.0µs
2.59×
Reported · license unknown · source
2026-03-14stale

Measured exactly what you asked. Reported by source; not independently reproduced.

source mirroredlicense unknownno install recipeView source →Run detail →
15
submission 553928bloomberg9383
26.1µs
2.59×
Reported · license unknown · source
2026-03-14stale

Measured exactly what you asked. Reported by source; not independently reproduced.

source mirroredlicense unknownno install recipeView source →Run detail →
16
submission 554356fluudgate
26.9µs
2.68×
Reported · license unknown · source
2026-03-14stale

Measured exactly what you asked. Reported by source; not independently reproduced.

source mirroredlicense unknownno install recipeView source →Run detail →
17
submission 552559lacalculatrice
27.8µs
2.77×
Reported · license unknown · source
2026-03-14stale

Measured exactly what you asked. Reported by source; not independently reproduced.

source mirroredlicense unknownno install recipeView source →Run detail →
18
submission 554994ramizzik
27.9µs
2.78×
Reported · license unknown · source
2026-03-14stale

Measured exactly what you asked. Reported by source; not independently reproduced.

source mirroredlicense unknownno install recipeView source →Run detail →
19
submission 553269ramizzik
27.9µs
2.78×
Reported · license unknown · no source
2026-03-14stale

Measured exactly what you asked. Reported by source; not independently reproduced.

no sourcelicense unknownno install recipeRun detail →
20
submission 554527jiannanWang
28.6µs
2.85×
Reported · license unknown · source
2026-03-14stale

Measured exactly what you asked. Reported by source; not independently reproduced.

source mirroredlicense unknownno install recipeView source →Run detail →
21
submission 555575happy_sloth_
29.4µs
2.92×
Reported · license unknown · source
2026-03-15stale

Measured exactly what you asked. Reported by source; not independently reproduced.

source mirroredlicense unknownno install recipeView source →Run detail →
22
submission 553113Narain
30.2µs
3.01×
Reported · license unknown · source
2026-03-14stale

Measured exactly what you asked. Reported by source; not independently reproduced.

source mirroredlicense unknownno install recipeView source →Run detail →
23
submission 554392kitrak_rev.
31.1µs
3.09×
Reported · license unknown · source
2026-03-14stale

Measured exactly what you asked. Reported by source; not independently reproduced.

source mirroredlicense unknownno install recipeView source →Run detail →
24
submission 553187brandonin
35.1µs
3.50×
Reported · license unknown · source
2026-03-14stale

Measured exactly what you asked. Reported by source; not independently reproduced.

source mirroredlicense unknownno install recipeView source →Run detail →
25
submission 553891suvasis_29047
39.0µs
3.88×
Reported · license unknown · source
2026-03-14stale

Measured exactly what you asked. Reported by source; not independently reproduced.

source mirroredlicense unknownno install recipeView source →Run detail →
26
submission 555585arapa.kr
40.9µs
4.07×
Reported · license unknown · source
2026-03-15stale

Measured exactly what you asked. Reported by source; not independently reproduced.

source mirroredlicense unknownno install recipeView source →Run detail →
27
submission 555433tr00ths3rum
47.2µs
4.69×
Reported · license unknown · source
2026-03-15stale

Measured exactly what you asked. Reported by source; not independently reproduced.

source mirroredlicense unknownno install recipeView source →Run detail →
28
submission 553118ramizzik
61.4µs
6.11×
Reported · license unknown · no source
2026-03-14stale

Measured exactly what you asked. Reported by source; not independently reproduced.

no sourcelicense unknownno install recipeRun detail →
29
submission 552686ramizzik
61.4µs
6.11×
Reported · license unknown · no source
2026-03-14stale

Measured exactly what you asked. Reported by source; not independently reproduced.

no sourcelicense unknownno install recipeRun detail →
30
submission 553238svdrecbd
62.1µs
6.18×
Reported · license unknown · source
2026-03-14stale

Measured exactly what you asked. Reported by source; not independently reproduced.

source mirroredlicense unknownno install recipeView source →Run detail →
31
submission 555564pradeep03071
64.2µs
6.39×
Reported · license unknown · source
2026-03-15stale

Measured exactly what you asked. Reported by source; not independently reproduced.

source mirroredlicense unknownno install recipeView source →Run detail →
32
submission 553072suvasis_29047
67.5µs
6.71×
Reported · license unknown · no source
2026-03-14stale

Measured exactly what you asked. Reported by source; not independently reproduced.

no sourcelicense unknownno install recipeRun detail →
33
submission 551097Seraphim
90.1µs
8.97×
Reported · license unknown · source
2026-03-14stale

Measured exactly what you asked. Reported by source; not independently reproduced.

source mirroredlicense unknownno install recipeView source →Run detail →
34
submission 554278dpang
106.3µs
10.58×
Reported · license unknown · no source
2026-03-14stale

Measured exactly what you asked. Reported by source; not independently reproduced.

no sourcelicense unknownno install recipeRun detail →
35
submission 552024yf225
373.7µs
37.19×
Reported · license unknown · source
2026-03-14stale

Measured exactly what you asked. Reported by source; not independently reproduced.

source mirroredlicense unknownno install recipeView source →Run detail →
36
submission 552578bloomberg9383
375.2µs
37.34×
Reported · license unknown · no source
2026-03-14stale

Measured exactly what you asked. Reported by source; not independently reproduced.

no sourcelicense unknownno install recipeRun detail →

Implementations

Implementation
Runtime
Best latency
Evidence
Availability
submission 555004brandonin
python
10.0µs
1.00×
Reported
license unknown · source
submission 555480Ayush10
python
12.6µs
1.25×
Reported
license unknown · source
submission 553892Mistobaan
python
13.6µs
1.35×
Reported
license unknown · source
submission 554383CodingMaster
python
14.0µs
1.39×
Reported
license unknown · source
submission 554374Narain
python
15.0µs
1.49×
Reported
license unknown · source
submission 554984kitrak_rev.
python
17.3µs
1.72×
Reported
license unknown · source
submission 555604the.defenestrator
python
19.9µs
1.98×
Reported
license unknown · source
submission 555492sissi1165
python
22.1µs
2.20×
Reported
license unknown · source
Show all 36 implementations ›
submission 555503alazarr.m
python
23.9µs
2.38×
Reported
license unknown · source
submission 555590fluudgate
python
25.1µs
2.50×
Reported
license unknown · source
submission 555174plops655
python
25.4µs
2.53×
Reported
license unknown · source
submission 555131dpang
python
26.0µs
2.58×
Reported
license unknown · source
submission 552586yhinai
python
26.0µs
2.59×
Reported
license unknown · source
submission 553928bloomberg9383
python
26.1µs
2.59×
Reported
license unknown · source
submission 552559lacalculatrice
python
27.8µs
2.77×
Reported
license unknown · source
submission 554994ramizzik
python
27.9µs
2.78×
Reported
license unknown · source
submission 554527jiannanWang
python
28.6µs
2.85×
Reported
license unknown · source
submission 555575happy_sloth_
python
29.4µs
2.92×
Reported
license unknown · source
submission 553891suvasis_29047
python
39.0µs
3.88×
Reported
license unknown · source
submission 555585arapa.kr
python
40.9µs
4.07×
Reported
license unknown · source
submission 555433tr00ths3rum
python
47.2µs
4.69×
Reported
license unknown · source
submission 553238svdrecbd
python
62.1µs
6.18×
Reported
license unknown · source
submission 555564pradeep03071
python
64.2µs
6.39×
Reported
license unknown · source
submission 551097Seraphim
python
90.1µs
8.97×
Reported
license unknown · source
submission 552024yf225
python
373.7µs
37.19×
Reported
license unknown · source
submission 553187brandonin
python
35.1µs
3.50×
Reported
license unknown · source
submission 554279Ayush10
python
13.9µs
1.38×
Reported
license unknown · source
submission 553113Narain
python
30.2µs
3.01×
Reported
license unknown · source
submission 554392kitrak_rev.
python
31.1µs
3.09×
Reported
license unknown · source
submission 554356fluudgate
python
26.9µs
2.68×
Reported
license unknown · source
submission 553269ramizzik
python
27.9µs
2.78×
Reported
license unknown · no source
submission 553118ramizzik
python
61.4µs
6.11×
Reported
license unknown · no source
submission 552686ramizzik
python
61.4µs
6.11×
Reported
license unknown · no source
submission 553072suvasis_29047
python
67.5µs
6.71×
Reported
license unknown · no source
submission 554278dpang
python
106.3µs
10.58×
Reported
license unknown · no source
submission 552578bloomberg9383
python
375.2µs
37.34×
Reported
license unknown · no source

Semantics

Inputs and outputs
xfp32 [b, d, s]
weightfp32 [d, w]
biasfp32 [d]
outfp32 [b, d, s]
Axes and behavior
bvariable
dvariable
svariable
wvariable
seedvariable
determinismunspecified
constraintsNo mutation or aliasing
Identity
aliascausal depthwise conv1daliascausal_conv1dsha256bfeae45ef25b…
Sources: GPU MODE KernelBot (2026-03-15) · June 9 Researcher Reciprocity License v1.0last observed 2026-03-15How records are decidedJSON