Skip to content
AI IntelligenceNov 14, 2026AI Intelligence
Article

Nvidia published a guide for serving GLM-5.3…

…a quantized 753-billion-parameter mixture-of-experts model from Z.ai with 40 billion parameters activated per token, across two DGX Stations connected via ConnectX-8 RDMA. The setup uses vLLM tensor parallelism to split weights and computation across both GPUs through a single endpoint, reaching 607.8 tokens per second.

AI-generated: summaries written by AI from the linked sources. How we use AI

AI-generatedSource: Hacker News
01

Source Brief

Nvidia published a guide for serving GLM-5.3, a quantized 753-billion-parameter mixture-of-experts model from Z.ai with 40 billion parameters activated per token, across two DGX Stations connected via ConnectX-8 RDMA. The setup uses vLLM tensor parallelism to split weights and computation across both GPUs through a single endpoint, reaching 607.8 tokens per second.