Compare runs
cohort sha256:2ecb90a5cb7f50b6… · source-native profile
ranking-v1
no ranks: not comparable
Comparison identity
operation
mla-ragged-prefill-causal-h16-qk192-vo128
workload
total_q = 86 · total_kv = 86 · len_indptr = 3 · bf16 · c93be321
protocol
flashinfer-bench · 7d44d206
environment
NVIDIA B200 · 83c9b6e7
correctness policy
64fb869e
metric
latency mean (ns)
Context · 10 fields›
architecture
sm_100
CUDA
13.0
driver
unknown
framework
pytorch 2.11.0+cu130
samples
unknown
evidence
reported
license
Apache-2.0
source
available
status
passed
observed
2026-04-06
No winner: these runs didn't measure the same thing. The rows above show what differs. Why comparable?
Add run ›
Export ›
Markdown CSV JSON
flashinfer_wrapper_d90b98 · sha256:7baa3e12a69a7… · observed 2026-04-06