NVIDIA L4
sm_89
Records held
8
Runs
131
Operations
8131 implementations
Records held on this GPU
Operation · workload
History
Record
Implementation
Since
Coverage by operation family
Family
Operations
Runs
With source
elementwise
2
38
33
reduction
1
26
26
conv
1
21
21
sort
1
13
13
gemm
1
12
12
scan
1
11
11
histogram
1
10
10
Serving evidence
End-to-end serving results measured on this hardware — a separate corpus with its own resolver, never ranked against the kernel records above.
Model · workload
Scenario
Best reported
Runs
Llama-3.1-8B (MLPerf reference)llama3.1-8b · Offline
offline
1,642 tok/s
1
Llama-3.1-8B (MLPerf reference)llama3.1-8b · Server
server
1,207 tok/s
1
2 serving runs · Open in the serving resolver →
Record activity
30Nov
10Dec
Jan
2Feb
9Mar
Apr
1May
Jun
Jul
Aug
Sep
Oct
Sources: GPU MODE KernelBot (2026-05-13) · June 9 Researcher Reciprocity License v1.0last observed 2026-05-13