Skip to content
KernelIndex
Search⌘K
Sign in

Proud Wren @ NVIDIA

Competition author · sol-execbench/user/Proud Wren @ NVIDIA · license unknown
Records held
3as of 2026-10-01 15:51 UTC
Entries
157
Runs
157
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
MoE expert batched execution with capacity factorsuite of 16 cases
2.14ms
4.7% faster · was 2.24 ms
Proud Wren @ NVIDIA
B200
2026-08-19
Language model decoder prenorm attention FFN residual backwardsuite of 16 cases
1.50ms
20.8% faster · was 1.89 ms
Proud Wren @ NVIDIA
B200
2026-07-12
KV shared attention with dual RoPEsuite of 16 cases
250.8µs
13.6% faster · was 290.2 µs
Proud Wren @ NVIDIA
B200
2026-07-08

Open in the records ledger →

Measured entries

Implementation
Operation
Runtime
Best latency
Evidence
Availability
Proud Wren @ NVIDIA
Multimodal RoPE position calculation
unknown
2.93µs
Reported
license unknown · no source
Proud Wren @ NVIDIA
Multimodal RoPE position calculation
unknown
2.96µs
Reported
license unknown · no source
Proud Wren @ NVIDIA
GEMM n4096 k4096
unknown
15.7µs
Reported
license unknown · no source
Proud Wren @ NVIDIA
MoE sparse routing and dispatch backward
unknown
20.7µs
Reported
license unknown · no source
Proud Wren @ NVIDIA
MoE sparse routing and dispatch backward
unknown
20.8µs
Reported
license unknown · no source
Proud Wren @ NVIDIA
Group limited topk routing
unknown
27.6µs
Reported
license unknown · no source
Proud Wren @ NVIDIA
Group limited topk routing
unknown
27.9µs
Reported
license unknown · no source
Proud Wren @ NVIDIA
Flux timestep guidance projection embedding
unknown
32.5µs
Reported
license unknown · no source
Proud Wren @ NVIDIA
Flux timestep guidance projection embedding
unknown
36.6µs
Reported
license unknown · no source
Proud Wren @ NVIDIA
Mamba discretization and segsum
unknown
36.8µs
Reported
license unknown · no source
Show all 40 entries ›
Proud Wren @ NVIDIA
Mamba discretization and segsum
unknown
36.9µs
Reported
license unknown · no source
Proud Wren @ NVIDIA
Mamba chunk scan with segsum
unknown
57.6µs
Reported
license unknown · no source
Proud Wren @ NVIDIA
SAM hq mask decoder IoU hypernetwork fusion backward
unknown
60.7µs
Reported
license unknown · no source
Proud Wren @ NVIDIA
Chunk gated delta rule linear attention
unknown
65.2µs
Reported
license unknown · no source
Proud Wren @ NVIDIA
Mamba chunk scan with segsum
unknown
68.0µs
Reported
license unknown · no source
Proud Wren @ NVIDIA
Cu seqlens variable length vision attention
unknown
75.6µs
Reported
license unknown · no source
Proud Wren @ NVIDIA
Flux concatenated sequence processing with split
unknown
76.0µs
Reported
license unknown · no source
Proud Wren @ NVIDIA
Altup predict correction cycle backward
unknown
78.9µs
Reported
license unknown · no source
Proud Wren @ NVIDIA
Seqlen-finetuned-reconstructed Hyena gating and output projection
unknown
95.9µs
Reported
license unknown · no source
Proud Wren @ NVIDIA
Seqlen-finetuned-reconstructed Hyena complete forward block
unknown
101.2µs
Reported
license unknown · no source
Proud Wren @ NVIDIA
Multi head QKV projection with RoPE backward
unknown
103.0µs
Reported
license unknown · no source
Proud Wren @ NVIDIA
Multi head QKV projection with RoPE backward
unknown
105.1µs
Reported
license unknown · no source
Proud Wren @ NVIDIA
Audio encoder varlen attention with chunking backward
unknown
114.3µs
Reported
license unknown · no source
Proud Wren @ NVIDIA
Audio encoder to language model multimodal fusion
unknown
115.8µs
Reported
license unknown · no source
Proud Wren @ NVIDIA
ConvNeXtV2 block with GRN
unknown
129.3µs
Reported
license unknown · no source
Proud Wren @ NVIDIA
Audio relative position attention
unknown
131.8µs
Reported
license unknown · no source
Proud Wren @ NVIDIA
Conformer relative position attention
unknown
133.7µs
Reported
license unknown · no source
Proud Wren @ NVIDIA
ResNet block with time embedding
unknown
142.2µs
Reported
license unknown · no source
Proud Wren @ NVIDIA
ResNet block with time embedding
unknown
147.3µs
Reported
license unknown · no source
Proud Wren @ NVIDIA
Audio encoder to language model multimodal fusion
unknown
166.4µs
Reported
license unknown · no source
Proud Wren @ NVIDIA
MoE complete layer with shared expert backward
unknown
174.2µs
Reported
license unknown · no source
Proud Wren @ NVIDIA
KV shared attention with RoPE
unknown
174.3µs
Reported
license unknown · no source
Proud Wren @ NVIDIA
Audio encoder to language model multimodal fusion
unknown
176.9µs
Reported
license unknown · no source
Proud Wren @ NVIDIA
Cross attention text video conditioning backward
unknown
178.0µs
Reported
license unknown · no source
Proud Wren @ NVIDIA
Cross attention text video conditioning backward
unknown
178.5µs
Reported
license unknown · no source
Proud Wren @ NVIDIA
FFN GELU projection fused backward
unknown
180.5µs
Reported
license unknown · no source
Proud Wren @ NVIDIA
FFN GELU projection fused backward
unknown
184.5µs
Reported
license unknown · no source
Proud Wren @ NVIDIA
Encoder layer dual residual norm chain
unknown
185.5µs
Reported
license unknown · no source
Proud Wren @ NVIDIA
Encoder layer dual residual norm chain
unknown
192.0µs
Reported
license unknown · no source
Proud Wren @ NVIDIA
ConvNeXtV2 block with GRN
unknown
197.7µs
Reported
license unknown · no source

Showing 40 of 157; the full list is in the API dossier below.

Record activity

Nov
Dec
Jan
Feb
Mar
1Apr
2May
1Jun
7Jul
1Aug
Sep
Oct
Sources: NVIDIA SOL-ExecBench (2026-08-23)last observed 2026-08-23JSON