submission 612159
dannywillowliu-uchi · python · License unknown
Use it
Vendorable · source mirrored · license unknownView source →
No package. Vendor the mirrored source: 16 lines, June 9 Researcher Reciprocity License v1.0.
submission.py
curl "https://kernelindex.com/api/v1/implementations/kernelbot-conv2d-v2-612159?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
Reported · How evidence levels are derived →
Source and license
sourceavailable
revision digestsha256:eca9b40c8bdb7413c7e37d277205ca03bb930630d08311140b2a1269b5f04820
license declaredunknown
license concludedunknown
authorsdannywillowliu-uchi
imported2026-08-15
Kernel source
submission.py16 lines
import os
os.environ['CUBLAS_WORKSPACE_CONFIG'] = ':4096:8'
from task import input_t, output_t
import torch
import torch.nn.functional as F
torch.backends.cudnn.allow_tf32 = False
torch.backends.cuda.matmul.allow_tf32 = False
torch.backends.cudnn.benchmark = True
def custom_kernel(data: input_t) -> output_t:
input_tensor, kernel, output = data
# F.conv2d allocates a new tensor. Can we avoid copy?
# Try: return the F.conv2d result directly (ignore output tensor)
return F.conv2d(input_tensor, kernel, stride=1, padding=0)
Source code from GPU Mode and the KernelBot dataset · June 9 Researcher Reciprocity License v1.0
Changes from previous submission
Against this author's previous submission submission 612044.
⋯ 9 unchanged linesdef custom_kernel(data: input_t) -> output_t:input_tensor, kernel, output = data- output[...] = F.conv2d(input_tensor, kernel, stride=1, padding=0)- return output+ # F.conv2d allocates a new tensor. Can we avoid copy?+ # Try: return the F.conv2d result directly (ignore output tensor)+ return F.conv2d(input_tensor, kernel, stride=1, padding=0)
Best evidence level for this revision: reported
JSON