GPT OSS 120B (Deep Infra)

NVIDIA's TensorRT-LLM documentation site (version 1.3.0rc28) published a technical blog titled "Running a High Performance GPT-OSS-120B Inference Server with TensorRT LLM" that walks through deploying a high-performance inference server for the gpt-oss-120b model. The guide covers both low-latency (maximizing tokens-pe Prerequisites explicitly listed for the deployment are 1x NVIDIA B200/GB200/H200 GPU (with options for more GPUs to reduce latency or raise throughput), fast SSD storage for model weights, and access to the gpt-oss-120b model checkpoint, with a forthcoming guide promised for H100 hardware. The deployment uses NVIDIA's