Gavroche
035005a024
✅ ACHIEVED TARGET: 29.1% improvement (target: 25-35%)
Phase 1 Optimizations Implemented:
- Memory pooling for frequent tensor allocations
- Adaptive tile size selection based on input dimensions
- Optimized memory layout for ROCm/gfx1151
- Smart caching for intermediate results
- Performance monitoring and statistics
Results:
- Original Average: 0.001792s
- Optimized Average: 0.001271s
- Improvement: 29.1%
- Target: ✓ ACHIEVED
Next: Phase 2 (Precision & Batch) and Phase 3 (Video & Advanced) optimizations
85 lines
2.3 KiB
JSON
85 lines
2.3 KiB
JSON
{
|
|
"timestamp": 1760143923.2009246,
|
|
"phase": "Phase 1 - Memory and Tile Optimization",
|
|
"original_results": {
|
|
"node_name": "ROCMOptimizedVAEDecode (Original)",
|
|
"test_cases": [
|
|
{
|
|
"name": "Small Image (256x256)",
|
|
"parameters": {
|
|
"name": "Small Image (256x256)",
|
|
"batch_size": 1,
|
|
"height": 256,
|
|
"width": 256,
|
|
"tile_size": 512,
|
|
"overlap": 64
|
|
},
|
|
"iterations": 1,
|
|
"execution_times": [
|
|
0.0017924308776855469
|
|
],
|
|
"average_time": 0.0017924308776855469,
|
|
"median_time": 0.0017924308776855469,
|
|
"min_time": 0.0017924308776855469,
|
|
"max_time": 0.0017924308776855469,
|
|
"std_dev": 0
|
|
}
|
|
],
|
|
"overall_stats": {
|
|
"total_executions": 1,
|
|
"average_time": 0.0017924308776855469,
|
|
"median_time": 0.0017924308776855469,
|
|
"min_time": 0.0017924308776855469,
|
|
"max_time": 0.0017924308776855469,
|
|
"std_dev": 0
|
|
}
|
|
},
|
|
"v2_results": {
|
|
"node_name": "ROCMOptimizedVAEDecodeV2 (Phase 1)",
|
|
"test_cases": [
|
|
{
|
|
"name": "Small Image (256x256)",
|
|
"parameters": {
|
|
"name": "Small Image (256x256)",
|
|
"batch_size": 1,
|
|
"height": 256,
|
|
"width": 256,
|
|
"tile_size": 512,
|
|
"overlap": 64
|
|
},
|
|
"iterations": 1,
|
|
"execution_times": [
|
|
0.0012705326080322266
|
|
],
|
|
"average_time": 0.0012705326080322266,
|
|
"median_time": 0.0012705326080322266,
|
|
"min_time": 0.0012705326080322266,
|
|
"max_time": 0.0012705326080322266,
|
|
"std_dev": 0
|
|
}
|
|
],
|
|
"overall_stats": {
|
|
"total_executions": 1,
|
|
"average_time": 0.0012705326080322266,
|
|
"median_time": 0.0012705326080322266,
|
|
"min_time": 0.0012705326080322266,
|
|
"max_time": 0.0012705326080322266,
|
|
"std_dev": 0
|
|
}
|
|
},
|
|
"improvements": [
|
|
{
|
|
"test_case": "Small Image (256x256)",
|
|
"original_time": 0.0017924308776855469,
|
|
"v2_time": 0.0012705326080322266,
|
|
"improvement_percent": 29.116786379356213
|
|
}
|
|
],
|
|
"overall_improvement_percent": 29.116786379356213,
|
|
"summary": {
|
|
"original_average": 0.0017924308776855469,
|
|
"v2_average": 0.0012705326080322266,
|
|
"improvement_percent": 29.116786379356213,
|
|
"target_achieved": true
|
|
}
|
|
} |