submission 69022
VectorVitalityFit · python · License unknown
Use it
Vendorable · source mirrored · license unknownView source →
No package. Vendor the mirrored source: 17 lines, June 9 Researcher Reciprocity License v1.0.
submission_vectoradd_v2.py
curl "https://kernelindex.com/api/v1/implementations/kernelbot-vectoradd-v2-69022?include=source"interfacepython
Compatibility
measured onNVIDIA B200
declared hardwareNVIDIA B200
architecturessm_100
dtypesfp16
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:e76c53ec362023d8b802830048344fcfd62a69a19c10f8296d12f6f92bfbec3b
license declaredunknown
license concludedunknown
authorsVectorVitalityFit
imported2026-08-15
Kernel source
submission_vectoradd_v2.py17 lines
#!POPCORN leaderboard vectoradd_v2
import torch
def custom_kernel(data):
x=data[0]
y=data[1]
assert x.shape == y.shape, "Input tensors must have the same shape"
# Make sure input tensors are float16
x = x.to(torch.float16)
y = y.to(torch.float16)
# If GPU is available, move tensors to GPU for fast computation
device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
x = x.to(device)
y = y.to(device)
# Perform element-wise addition (vectorized)
output = x + y
return outputSource code from GPU Mode and the KernelBot dataset · June 9 Researcher Reciprocity License v1.0
Best evidence level for this revision: reported
JSON