Unverified50% confidenceFactExact time
DeepSeek使用的H800在NVLink带宽上相比H100存在削减
1
Sources
50%
Confidence
Long-term
Relevance
7/12/2026
First Seen
Sources
DeepSeek Builds Its Own AI Chips: The Strategic Logic and Challenges Behind Compute Independence
hackernewshackernews7/9/2026
Related Claims
UnverifiedDeepSeek早期模型使用英伟达H800芯片训练完成，H800是英伟达专为中国市场削减互联带宽后推出的H100替代版本71% similarUnverifiedFull deployment of DeepSeek-V3.2 typically requires a cluster of multiple high-end GPUs (such as NVIDIA A100/H100)65% similarUnverifiedDeepStream supports deployment on data centers equipped with A100/H100, processing hundreds of video streams through horizontal scaling.64% similarUnverifiedDeepSeek's official servers have long been under high load, users frequently encounter 'server busy' messages, and access to the full R1 model across all versions is extremely difficult.63% similarUnverifiedDeepSeek V4 is said to have used an improved version of Hyper-Connections and achieved good results.62% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/491027API
curl https://kongchang.com/api/v1/knowledge/claims/491027MCP
get_claim(id=491027)