Skip to content
KernelIndex
Search⌘K

Flux concatenated sequence processing with split

20 eligible runs
diffusion

Flux concatenated sequence processing pattern: concatenates text and image sequences, applies linear projection, then splits back. This pattern appears in dual-stream transformer blocks and represents a memory bandwidth bottleneck.

Fastest reported
65.5µsmean · 68% of SOL
RIAC as wellLicense unknown · unknown

Reported evidence · last observed 2026-08-18. Reported by source; not independently reproduced. No public source. License unknown.

No public sourceRun detail →

Current records

Not measured on H100 for this workload. Challenges →

Source-native comparison · GPU NVIDIA B200 · Workload suite of 16 cases · mean latency · Protocol SOL-ExecBench evaluation stack v1.1 · mean · 14 results · last observed 2026-08-18Record history →
#
Implementation
Latency
vs #1
Trust
Observed
1
65.5µs68% SOL
1.00×
Reported · license unknown · no source
2026-08-18

Measured exactly what you asked. Reported by source; not independently reproduced.

no sourcelicense unknownno install recipeRun detail →
2
66.9µs66% SOL
1.02×
Reported · license unknown · no source
2026-07-10

Measured exactly what you asked. Reported by source; not independently reproduced.

no sourcelicense unknownno install recipeRun detail →
3
69.0µs65% SOL
1.05×
Reported · license unknown · no source
2026-08-15

Measured exactly what you asked. Reported by source; not independently reproduced.

no sourcelicense unknownno install recipeRun detail →
4
70.1µs65% SOL
1.07×
Reported · license unknown · no source
2026-08-14

Measured exactly what you asked. Reported by source; not independently reproduced.

no sourcelicense unknownno install recipeRun detail →
5
71.0µs65% SOL
1.08×
Reported · license unknown · no source
2026-08-04

Measured exactly what you asked. Reported by source; not independently reproduced.

no sourcelicense unknownno install recipeRun detail →
6
73.2µs63% SOL
1.12×
Reported · license unknown · no source
2026-07-17

Measured exactly what you asked. Reported by source; not independently reproduced.

no sourcelicense unknownno install recipeRun detail →
7
74.9µs63% SOL
1.14×
Reported · license unknown · no source
2026-07-19

Measured exactly what you asked. Reported by source; not independently reproduced.

no sourcelicense unknownno install recipeRun detail →
8
75.1µs62% SOL
1.15×
Reported · license unknown · no source
2026-07-10

Measured exactly what you asked. Reported by source; not independently reproduced.

no sourcelicense unknownno install recipeRun detail →
9
76.0µs62% SOL
1.16×
Reported · license unknown · no source
2026-07-11

Measured exactly what you asked. Reported by source; not independently reproduced.

no sourcelicense unknownno install recipeRun detail →
10
76.4µs62% SOL
1.17×
Reported · license unknown · no source
2026-08-15

Measured exactly what you asked. Reported by source; not independently reproduced.

no sourcelicense unknownno install recipeRun detail →
11
90.3µs55% SOL
1.38×
Reported · license unknown · no source
2026-07-27

Measured exactly what you asked. Reported by source; not independently reproduced.

no sourcelicense unknownno install recipeRun detail →
12
100.9µs51% SOL
1.54×
Reported · license unknown · no source
2026-08-13

Measured exactly what you asked. Reported by source; not independently reproduced.

no sourcelicense unknownno install recipeRun detail →
13
112.2µs48% SOL
1.71×
Reported · license unknown · no source
2026-07-10

Measured exactly what you asked. Reported by source; not independently reproduced.

no sourcelicense unknownno install recipeRun detail →
14
237.2µs29% SOL
3.62×
Reported · license unknown · no source
2026-07-10

Measured exactly what you asked. Reported by source; not independently reproduced.

no sourcelicense unknownno install recipeRun detail →

Implementations

Implementation
Runtime
Best latency
Evidence
Availability
unknown
65.5µs
1.04×
Reported
license unknown · no source
unknown
71.0µs
1.13×
Reported
license unknown · no source
unknown
73.2µs
1.16×
Reported
license unknown · no source
unknown
74.9µs
1.19×
Reported
license unknown · no source
unknown
76.0µs
1.20×
Reported
license unknown · no source
unknown
76.4µs
1.21×
Reported
license unknown · no source
unknown
107.4µs
1.70×
Reported
license unknown · no source
unknown
69.0µs
1.09×
Reported
license unknown · no source
Show all 20 implementations ›
unknown
70.1µs
1.11×
Reported
license unknown · no source
unknown
75.1µs
1.19×
Reported
license unknown · no source
unknown
81.2µs
1.29×
Reported
license unknown · no source
unknown
90.3µs
1.43×
Reported
license unknown · no source
unknown
100.9µs
1.60×
Reported
license unknown · no source
unknown
112.2µs
1.78×
Reported
license unknown · no source
unknown
237.2µs
3.76×
Reported
license unknown · no source
unknown
260.5µs
4.13×
Reported
license unknown · no source
unknown
377.3µs
5.98×
Reported
license unknown · no source
unknown
63.1µs
1.00×
Reported
license unknown · no source
unknown
66.9µs
1.06×
Reported
license unknown · no source
unknown
68.7µs
1.09×
Reported
license unknown · no source

Semantics

Inputs and outputs
hidden_statesfp32 [batch_size, img_seq_len, hidden_dim]
encoder_hidden_statesfp32 [batch_size, text_seq_len, hidden_dim]
process_weightfp32 [hidden_dim, hidden_dim]
processed_encoderfp32 [batch_size, text_seq_len, hidden_dim]
processed_hiddenfp32 [batch_size, img_seq_len, hidden_dim]
Axes and behavior
batch_sizevariable
hidden_dimconstant = 3072
img_seq_lenvariable
text_seq_lenvariable
determinismunspecified
constraintsNo mutation or aliasing
Identity
alias030_flux_concatenated_sequence_processing_with_splitmodelflux-1-kontext-devsha2567a7c32168bab…
Sources: NVIDIA SOL-ExecBench (2026-08-18)last observed 2026-08-18How records are decidedJSON