Skip to content
KernelIndex
Search⌘K
Sign in

Gleaming Coyote

Competition author · sol-execbench/user/Gleaming Coyote · license unknown
Records held
1as of 2026-10-02 01:32 UTC
Entries
45
Runs
45
Is this you? Claim this project ›

A claim grants attribution and metadata maintenance, never the right to edit evidence. Evidence claims are reviewed by a maintainer.

Current records

Operation · workload
History
Record
Implementation
Hardware
Since
Text decoder layer with self attention and MLPsuite of 16 cases
669.2µs
4.5% faster · was 700.6 µs
Gleaming Coyote
B200
2026-08-14

Open in the records ledger →

Measured entries

Implementation
Operation
Runtime
Best latency
Evidence
Availability
Gleaming Coyote
FP8 YaRN RoPE embedding
unknown
3.40µs
Reported
license unknown · no source
Gleaming Coyote
RMSNorm h128
unknown
3.71µs
Reported
license unknown · no source
Gleaming Coyote
RMSNorm h2048
unknown
4.14µs
Reported
license unknown · no source
Gleaming Coyote
RMSNorm h4096
unknown
6.12µs
Reported
license unknown · no source
Gleaming Coyote
Fused add RMSNorm h7168
unknown
7.23µs
Reported
license unknown · no source
Gleaming Coyote
FP8 MoE router projection
unknown
11.4µs
Reported
license unknown · no source
Gleaming Coyote
Attention score value matmul
unknown
15.2µs
Reported
license unknown · no source
Gleaming Coyote
FP8 Mamba2 SSM discretization
unknown
16.6µs
Reported
license unknown · no source
Gleaming Coyote
Group limited topk routing
unknown
21.6µs
Reported
license unknown · no source
Gleaming Coyote
Grouped topk MoE routing backward
unknown
25.3µs
Reported
license unknown · no source
Show all 40 entries ›
Gleaming Coyote
Flux multi head RMSNorm QK
unknown
25.3µs
Reported
license unknown · no source
Gleaming Coyote
Seqlen-finetuned-reconstructed Hyena complete forward block
unknown
26.3µs
Reported
license unknown · no source
Gleaming Coyote
Attention QKV projection with GQA reshape
unknown
30.3µs
Reported
license unknown · no source
Gleaming Coyote
Attention QKV with QK norm single kernel backward
unknown
34.8µs
Reported
license unknown · no source
Gleaming Coyote
Audio attention QKV projection with normalization
unknown
47.3µs
Reported
license unknown · no source
Gleaming Coyote
FP8 MoE gate routing
unknown
51.1µs
Reported
license unknown · no source
Gleaming Coyote
Fused gated MLP SiLU
unknown
97.9µs
Reported
license unknown · no source
Gleaming Coyote
Mamba chunk scan with segsum
unknown
99.1µs
Reported
license unknown · no source
Gleaming Coyote
Grouped query self attention with RoPE
unknown
116.5µs
Reported
license unknown · no source
Gleaming Coyote
NVFP4 MoE expert linear with gating
unknown
118.0µs
Reported
license unknown · no source
Gleaming Coyote
Audio encoder varlen attention with chunking backward
unknown
119.6µs
Reported
license unknown · no source
Gleaming Coyote
Joint attention context projection
unknown
146.9µs
Reported
license unknown · no source
Gleaming Coyote
FFN GELU projection fused backward
unknown
150.2µs
Reported
license unknown · no source
Gleaming Coyote
Fused attention QK matmul scale mask softmax backward
unknown
155.0µs
Reported
license unknown · no source
Gleaming Coyote
MoE complete layer with shared expert backward
unknown
163.6µs
Reported
license unknown · no source
Gleaming Coyote
MLA fused QKV RoPE split
unknown
172.2µs
Reported
license unknown · no source
Gleaming Coyote
FP8 Mamba2 in projection
unknown
182.4µs
Reported
license unknown · no source
Gleaming Coyote
GIT vision encoder layer
unknown
218.8µs
Reported
license unknown · no source
Gleaming Coyote
ConvNeXtV2 layer with NHWC persistence backward
unknown
219.2µs
Reported
license unknown · no source
Gleaming Coyote
NVFP4 fused MLP SiLU gating
unknown
236.5µs
Reported
license unknown · no source
Gleaming Coyote
Grouped query attention with RoPE and QK norm
unknown
256.4µs
Reported
license unknown · no source
Gleaming Coyote
Fused gate up projection with SwiGLU
unknown
333.9µs
Reported
license unknown · no source
Gleaming Coyote
Dual stream attention with conditional cross attention
unknown
346.5µs
Reported
license unknown · no source
Gleaming Coyote
Decoder complete layer
unknown
398.1µs
Reported
license unknown · no source
Gleaming Coyote
Feedforward MLP backward
unknown
433.0µs
Reported
license unknown · no source
Gleaming Coyote
MoE expert computation with weighted accumulation
unknown
507.3µs
Reported
license unknown · no source
Gleaming Coyote
Text decoder layer with self attention and MLP
unknown
669.2µs
Reported
license unknown · no source
Gleaming Coyote
Basic transformer block
unknown
804.5µs
Reported
license unknown · no source
Gleaming Coyote
MoE expert MLP with load balancing
unknown
1.00ms
Reported
license unknown · no source
Gleaming Coyote
GQA rotary attention core backward
unknown
1.50ms
Reported
license unknown · no source

Showing 40 of 45; the full list is in the API dossier below.

Record activity

Nov
Dec
Jan
Feb
Mar
Apr
May
Jun
1Jul
4Aug
Sep
Oct
Sources: NVIDIA SOL-ExecBench (2026-08-15)last observed 2026-08-15JSON