Entry #4
Completed- Hotkey
- 5HQdTR…iUhTk5
- Sent
- 10 Nov 2026, 18:31 UTC
- Source
- 5d61fb76fc88
Progress
- Queued
- Proof
- Planned
- Timing
- Done
Done means every board has a decision. It does not mean the entry took a board over.
On each board
| Board | Runs | Time | Decision |
|---|---|---|---|
| Overallvs #8 | Done2 of 2 runs | 5.041×Generated 5.421×, real models 4.688× | Above the barGain lower bound -436.5%Guard passed |
| MLPvs #1 | Done2 of 2 runs | 46.753×Generated 53.904×, real models 40.550× | Above the barGain lower bound -5799.8%Guard passed |
| ResNetvs #9 | Done2 of 2 runs | 1.981×Generated 3.270×, real models 1.200× | Above the barGain lower bound -145.6%Guard passed |
| MobileNetvs the shipped reference | Done2 of 2 runs | 1.001×Generated 1.001×, real models 1.000× | Above the barGain lower bound -0.3%Guard passed |
| Normalized MLPvs #6 | Done2 of 2 runs | 65.269×Generated 85.167×, real models 50.019× | Above the barGain lower bound -7803.9%Guard passed |
| Mixervs #7 | Done2 of 2 runs | 105.881×Generated 138.585×, real models 80.895× | Above the barGain lower bound -12648.6%Guard passed |
| Recurrentvs the shipped reference | Done2 of 2 runs | 0.528×Generated 0.536×, real models 0.521× | QualifiesGain lower bound 45.3%Guard passed |
Every run
Each board times the entry on every benchmark machine, over fresh seeds, on generated workloads and on trained models with real data. A run's time combines the two suites by their published weights; the bars below split it by model, batch and input, and by real model.
Overall2 runs
RunsGenerated 5.421× · real models 4.688×
MachineTime
- prod-amd64Seed 10097558775.271×Done1 attempt0.804× of the baseline kernels' time
- prod-arm64Seed 10097558774.822×Done1 attempt0.812× of the baseline kernels' time
By model and inputGuard at 1.100×
MLP
- Batch 1, normal54.631×
- Batch 1, sparse41.895×
- Batch 1, wide50.670×
Mixer
- Batch 1, normal127.652×
- Batch 1, sparse129.592×
- Batch 1, wide135.746×
ResNet
- Batch 1, normal1.037×
- Batch 1, sparse1.023×
- Batch 1, wide1.036×
MobileNet
- Batch 1, normal0.840×
- Batch 1, sparse0.824×
- Batch 1, wide0.840×
Recurrent
- Batch 1, normal0.060×
- Batch 1, sparse0.065×
- Batch 1, wide0.061×
Normalized MLP
- Batch 1, normal75.675×
- Batch 1, sparse63.837×
- Batch 1, wide76.353×
Real models
- appliance-energy-mixer on Appliance Energy Prediction197.019×
- compact-mixer on digits70.463×
- compact-mixer on fashion104.398×
- compact-mixer on mnist27.988×
- compact-mlp on digits22.632×
- compact-mlp on fashion41.945×
- compact-mlp on mnist25.341×
- compact-mobilenet on digits0.832×
- compact-mobilenet on fashion0.813×
- compact-mobilenet on mnist0.843×
- compact-normalized_mlp on digits76.407×
- compact-normalized_mlp on fashion75.934×
- compact-normalized_mlp on mnist58.630×
- compact-resnet on digits1.034×
- compact-resnet on fashion1.034×
- compact-resnet on mnist1.034×
- appliance-gru on appliance-energy0.060×
- mitbih-gru on mitbih0.052×
- mnist-rows-lstm on mnist-rows0.055×
- speech-commands-gru on speech-commands0.071×
- speech-commands-lstm-gru on speech-commands0.062×
- spoken-arabic-digit-lstm on spoken-arabic-digit0.050×
- tiny-shakespeare-lstm on tiny-shakespeare0.066×
- uci-har-gru on uci-har0.057×
- uci-har-lstm on uci-har0.051×
- uci-har-mixer on UCI-HAR126.642×
- uci-har-mlp on UCI-HAR76.240×
- uci-har-mobilenet on UCI-HAR0.823×
- uci-har-normalized_mlp on UCI-HAR68.641×
- uci-har-resnet on UCI-HAR0.872×
MLP2 runs
RunsGenerated 53.904× · real models 40.550×
MachineTime
- prod-amd64Seed 100975587756.295×Done1 attempt1.002× of the baseline kernels' time
- prod-arm64Seed 100975587738.828×Done1 attempt1.005× of the baseline kernels' time
By model and inputGuard at 1.100×
MLP
- Batch 1, normal59.499×
- Batch 1, sparse47.633×
- Batch 1, wide59.814×
Real models
- compact-mlp on digits20.199×
- compact-mlp on fashion51.030×
- compact-mlp on mnist36.757×
- uci-har-mlp on UCI-HAR71.359×
ResNet2 runs
RunsGenerated 3.270× · real models 1.200×
MachineTime
- prod-amd64Seed 10097558772.047×Done1 attempt1.008× of the baseline kernels' time
- prod-arm64Seed 10097558771.916×Done1 attempt0.999× of the baseline kernels' time
By model and inputGuard at 1.100×
ResNet
- Batch 1, normal3.308×
- Batch 1, sparse3.228×
- Batch 1, wide3.356×
Real models
- compact-resnet on digits1.745×
- compact-resnet on fashion0.913×
- compact-resnet on mnist0.432×
- uci-har-resnet on UCI-HAR3.011×
MobileNet2 runs
RunsGenerated 1.001× · real models 1.000×
MachineTime
- prod-amd64Seed 10097558771.004×Done1 attempt1.003× of the baseline kernels' time
- prod-arm64Seed 10097558770.997×Done1 attempt0.998× of the baseline kernels' time
By model and inputGuard at 1.100×
MobileNet
- Batch 1, normal1.000×
- Batch 1, sparse1.001×
- Batch 1, wide1.002×
Real models
- compact-mobilenet on digits1.000×
- compact-mobilenet on fashion0.999×
- compact-mobilenet on mnist1.000×
- uci-har-mobilenet on UCI-HAR1.001×
Normalized MLP2 runs
RunsGenerated 85.167× · real models 50.019×
MachineTime
- prod-amd64Seed 100975587767.576×Done1 attempt0.771× of the baseline kernels' time
- prod-arm64Seed 100975587763.040×Done1 attempt0.764× of the baseline kernels' time
By model and inputGuard at 1.100×
Normalized MLP
- Batch 1, normal82.673×
- Batch 1, sparse79.646×
- Batch 1, wide89.588×
Real models
- compact-normalized_mlp on digits73.146×
- compact-normalized_mlp on fashion42.033×
- compact-normalized_mlp on mnist36.599×
- uci-har-normalized_mlp on UCI-HAR55.629×
Mixer2 runs
RunsGenerated 138.585× · real models 80.895×
MachineTime
- prod-amd64Seed 1009755877108.027×Done1 attempt0.686× of the baseline kernels' time
- prod-arm64Seed 1009755877103.777×Done1 attempt0.692× of the baseline kernels' time
By model and inputGuard at 1.100×
Mixer
- Batch 1, normal141.618×
- Batch 1, sparse138.177×
- Batch 1, wide144.405×
Real models
- appliance-energy-mixer on Appliance Energy Prediction187.253×
- compact-mixer on digits55.643×
- compact-mixer on fashion100.383×
- compact-mixer on mnist25.246×
- uci-har-mixer on UCI-HAR131.193×
Recurrent2 runs
RunsGenerated 0.536× · real models 0.521×
MachineTime
- prod-amd64Seed 10097558770.532×Done1 attempt0.540× of the baseline kernels' time
- prod-arm64Seed 10097558770.524×Done1 attempt0.532× of the baseline kernels' time
By model and inputGuard at 1.100×
Recurrent
- Batch 1, normal0.530×
- Batch 1, sparse0.575×
- Batch 1, wide0.523×
Real models
- appliance-gru on appliance-energy0.517×
- mitbih-gru on mitbih0.529×
- mnist-rows-lstm on mnist-rows0.531×
- speech-commands-gru on speech-commands0.488×
- speech-commands-lstm-gru on speech-commands0.499×
- spoken-arabic-digit-lstm on spoken-arabic-digit0.530×
- tiny-shakespeare-lstm on tiny-shakespeare0.525×
- uci-har-gru on uci-har0.516×
- uci-har-lstm on uci-har0.552×