Unverified50% confidenceFactExact time
一个2万亿参数的模型即便使用FP8量化，其权重也高达2TB，每生成一个token就需完整扫描这2TB数据
1
Sources
50%
Confidence
Long-term
Relevance
7/22/2026
First Seen
Sources
Related Claims
UnverifiedA model with 100 billion parameters requires approximately 80 GB of VRAM to process a single inference request, while a model with tens of billions of parameters requires only about 20 GB.69% similarUnverifiedThe GPU resources required for a single inference of a hundred-billion-parameter flagship model may be 10 to 20 times higher than those of a ten-billion-parameter model.69% similarUnverifiedAnalysis speculates that the model is an MoE model with 2 to 4 trillion parameters, activating only approximately 150 billion parameters per token.67% similarUnverifiedE2B后缀表示Effective 2B，模型总参数量更大但每次推理只激活约20亿有效参数66% similarUnverifiedGrok 1 的3140亿总参数中每次推理仅激活约25%，实际推理算力接近650亿参数规模的密集模型65% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/587229API
curl https://kongchang.com/api/v1/knowledge/claims/587229MCP
get_claim(id=587229)