单 GPU 不够用：用 vLLM 在两台 NVIDIA DGX Station 上部署 GLM-5.3
When One GPU Isn’t Enough: Serving GLM-5.3 Across Two NVIDIA DGX Stations with vLLM
NVIDIA 展示通过 ConnectX-8 RDMA 连接两台 DGX Station，用 vLLM 以张量并行把 GLM-5.3 的权重与计算拆分到两块 GPU 上，通过单一端点提供推理服务。所用模型为 NVIDIA GLM-5.3 NVFP4，是 Z.ai 混合专家模型的量化版本。
Running a frontier-scale model locally means keeping all its weights in GPU memory; the largest models no longer fit on one machine. Two NVIDIA DGX Stations...
Running a frontier-scale model locally means keeping all its weights in GPU memory; the largest models no longer fit on one machine. Two NVIDIA DGX Stations connected over ConnectX-8 RDMA serve one model through a single endpoint, with tensor parallelism splitting the model weights and computation across both GPUs. The NVIDIA GLM-5.3 NVFP4 model, a quantized version of Z.ai’s mixture-of…
来源：NVIDIA Technical Blog · developer.nvidia.com