feat: Runtime output buffer optimization #2285

Job	Run time
generate-matrix / generate	3s
filter-matrix	6s
Build torch-tensorrt whl package (pytorch/tensorrt, packaging/pre_build_script.sh, packaging/env_... / manywheel-py3_9-cuda11_8	7m 32s
Build torch-tensorrt whl package (pytorch/tensorrt, packaging/pre_build_script.sh, packaging/env_... / manywheel-py3_9-cuda12_4	7m 55s
Build torch-tensorrt whl package (pytorch/tensorrt, packaging/pre_build_script.sh, packaging/env_... / manywheel-py3_9-cuda12_6	14m 50s
Build torch-tensorrt whl package (pytorch/tensorrt, packaging/pre_build_script.sh, packaging/env_... / upload / manywheel-py3_9-cuda11_8	10s
Build torch-tensorrt whl package (pytorch/tensorrt, packaging/pre_build_script.sh, packaging/env_... / upload / manywheel-py3_9-cuda12_4	14s
Build torch-tensorrt whl package (pytorch/tensorrt, packaging/pre_build_script.sh, packaging/env_... / upload / manywheel-py3_9-cuda12_6	15s
Test torchscript frontend [Python]	0s
Test dynamo frontend [Python]	0s
Test dynamo core [Python]	0s
Test dynamo converters [Python]	0s
Test dynamo export serde [Python]	0s
Test torch compile backend [Python]	0s
Test dynamo cudagraphs [Python]	0s
Test core [Python]	0s
	31m 5s

Provide feedback