Skip to content
KernelIndex
Search⌘K

submission 183714

tomaszki · python · License unknown

Use it

Vendorable · source mirrored · license unknownView source →

No package. Vendor the mirrored source: 8 lines, June 9 Researcher Reciprocity License v1.0.

prefixsum.py
curl "https://kernelindex.com/api/v1/implementations/kernelbot-prefixsum-v2-183714?include=source"
interfacepython
Compatibility
measured onNVIDIA B200
declared hardwareNVIDIA B200
architecturessm_100
dtypesfp32

Benchmark evidence

1 measurement across 1 GPU, fastest first.

Operation / workload
Hardware
Latency
Rank
Observed
Inclusive prefix sumsuite of 11 cases
NVIDIA B200
638.8µs
#12 of 23
2025-12-20

Reported · How evidence levels are derived →

Source and license

sourceavailable
revision digestsha256:63581a3262d08d5e15bc58a9712148383a452af6f98d53f8ee5404794d6a7d60
license declaredunknown
license concludedunknown
authorstomaszki
imported2026-08-15

Kernel source

prefixsum.py8 lines
import torch
from task import input_t, output_t


def custom_kernel(data: input_t) -> output_t:
    data, output = data
    output = torch.cumsum(data, dim=0)
    return output

Source code from GPU Mode and the KernelBot dataset · June 9 Researcher Reciprocity License v1.0

Best evidence level for this revision: reported

JSON