Unverified50% confidenceFactExact time
The GLM-5.3 checkpoint in NVFP4 format is about 464 GB, while a single GB300 GPU has only 252 GB of memory, so the model cannot fit into a single GPU.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
11/14/2026
First Seen
Valid until: 2/12/2027
Sources
Related Entities
Related Claims
UnverifiedThe T4 GPU has 16GB of VRAM and 8.1 TFLOPS of FP32 compute performance.76% similarVerifiedA GPT-4-class model with 175 billion parameters requires approximately 175 GB of VRAM even when stored at fp8 precision.71% similarUnverifiedA 4B parameter model (such as Phi-3-mini or Llama 3.2 3B) requires only approximately 8GB of VRAM and can run on a consumer-grade GPU.71% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1170769API
curl https://kongchang.com/api/v1/knowledge/claims/1170769MCP
get_claim(id=1170769)