Unverified95% confidenceEventTime unknown
开发者使用4张3080Ti 16G魔改版显卡本地部署千问3.6 27B FP8模型进行编程实测
1
Sources
95%
Confidence
Medium-term (~90 days)
Relevance
5/31/2026
First Seen
Valid until: 8/29/2026
Sources
4×3080Ti本地部署千问3.6 27B跑OpenCode编程实测
bilibili苏不二师兄
Related Entities
Related Claims
UnverifiedQwen 3.6 27B FP8模型在4×3080Ti 16GB GPU上成功完成本地部署74% similarUnverified35B模型在FP16精度下需要约70GB显存，远超RTX 5090的32GB容量，因此单卡部署需量化71% similarUnverified混元3开源了BF16版本，并提供FP8量化版本，FP8版本约300GB70% similarUnverifiedT4 GPU拥有16GB显存和8.1 TFLOPS的FP32算力70% similarUnverified2.4万亿参数模型在FP32下约需9.6TB显存，BF16约需4.8TB，INT8约需2.4TB，INT4约需1.2TB70% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/4395API
curl https://kongchang.com/api/v1/knowledge/claims/4395MCP
get_claim(id=4395)