{"slug": "nvidia-s-local-glm-checkpoint-needs-two-gpus-to-run", "title": "NVIDIA's local GLM checkpoint needs two GPUs to run", "summary": "NVIDIA's FP4-Local checkpoint of Z.ai's GLM-5.3-Flash, released October 5th, requires two RTX PRO 6000 Blackwell GPUs or two DGX Spark systems in every configuration NVIDIA has validated, using about 166 GiB of VRAM. NVIDIA RTX Spark promoted the checkpoint on October 20th alongside a Qwen counterpart that occupies about 96 GiB on disk and is validated on a single DGX Spark or RTX PRO 6000 Blackwell GPU with offloading. The gap means running GLM-5.3-Flash locally still demands a multi-GPU or multi-system setup.", "body_md": "# NVIDIA's local GLM checkpoint needs two GPUs to run\n\n**Released October 5th, the FP4 version of Z.ai's GLM-5.3-Flash uses about 166 GiB of VRAM; Qwen's counterpart occupies about 96 GiB on disk.**\n\n        By [Ryan Merket](https://runtimewire.com/author/ryan-merket)\n        · Published \n\nPrimary source: [X](https://x.com/NVIDIARTXSpark/status/2112635102069248023)\n\n## Why it matters\n\nNVIDIA's Qwen checkpoint is validated on a single DGX Spark or RTX PRO 6000 Blackwell GPU with offloading, while GLM's listed configurations require two GPUs or two DGX Spark systems. Local inference can still require a multi-GPU or multi-system setup.\n\nNVIDIA's FP4-Local version of Z.ai's [GLM-5.3-Flash](https://runtimewire.com/models/z-ai/glm-5.3-flash:batch) requires two RTX PRO 6000 Blackwell GPUs or two DGX Spark systems in the configurations NVIDIA has validated. The checkpoint went up on October 5th; [NVIDIA RTX Spark's October 20th post](https://x.com/NVIDIARTXSpark/status/2112635102069248023) promoted it alongside a smaller Qwen checkpoint as a way to run large models locally.…\n\nWhen you purchase through links in our articles, [we may earn a small commission](https://runtimewire.com/affiliate-monetization-standards). This doesn’t affect our editorial independence.", "url": "https://wpnews.pro/news/nvidia-s-local-glm-checkpoint-needs-two-gpus-to-run", "canonical_source": "https://runtimewire.com/article/nvidia-fp4-local-qwen-glm-checkpoints", "published_at": "2026-10-20 20:42:34+00:00", "updated_at": "2026-10-20 23:48:44.496353+00:00", "lang": "en", "topics": ["large-language-models", "ai-infrastructure", "ai-chips", "ai-tools"], "entities": ["NVIDIA", "Z.ai", "GLM-5.3-Flash", "Qwen", "RTX PRO 6000 Blackwell", "DGX Spark", "NVIDIA RTX Spark"], "also_reported_by": [], "alternates": {"html": "https://wpnews.pro/news/nvidia-s-local-glm-checkpoint-needs-two-gpus-to-run", "markdown": "https://wpnews.pro/news/nvidia-s-local-glm-checkpoint-needs-two-gpus-to-run.md", "text": "https://wpnews.pro/news/nvidia-s-local-glm-checkpoint-needs-two-gpus-to-run.txt", "jsonld": "https://wpnews.pro/news/nvidia-s-local-glm-checkpoint-needs-two-gpus-to-run.jsonld"}}