[TRTLLM-16460][perf] Qwen3.8-Flash-Next NVFP4 modeling_v2 target: GB300 Pareto optimizations for ISL 8K / OSL 1K - #20937

[TRTLLM-16460][perf] Qwen3.8-Flash-Next NVFP4 modeling_v2 target: GB300 Pareto optimizations for ISL 8K / OSL 1K#20937

Wanli-Jiang wants to merge 5 commits into

NVIDIA:mainNVIDIA/TensorRT-LLM:mainfrom

Wanli-Jiang:user/williamj/qwen38fn-gb300-paretoWanli-Jiang/TensorRT-LLM:user/williamj/qwen38fn-gb300-paretoCopy head branch name to clipboard