Unverified50% confidenceBenchmarkExact time
On the DeepSeek V4 Pro workload, with an interaction target of 160 tokens/second per user, the per-megawatt AI factory throughput of Vera Rubin-NVL72 is up to 30 times higher than that of GB300-NVL72.
1
Sources
50%
Confidence
Short-term (~14 days)
Relevance
10/14/2026
First Seen
Valid until: 10/28/2026
Sources
Related Entities
Related Claims
Unverified使用NVIDIA B300芯片在1兆瓦数据中心可实现约45万token/s总吞吐量，若每用户340 token/s则约1324个并发活跃用户74% similarUnverifiedThe decode unit of DeepSeek V3 uses 144-card EP, amortizing the weight load overhead across large batches and freeing up GPU memory to increase batch size.73% similarUnverifiedIn the DeepSeek V1 era, each incremental token required nearly 400 KB of KV cache, which dropped to 48 KB in V3.2.73% similarUnverifiedDeepSeek-V3 has a total of 685B parameters, and even after Q4 quantization its size still exceeds 400GB.72% similarUnverified测试者用双4090（48GB显存）工作站运行20.7B稠密模型，每秒生成27.6个Token，仅比矮星略快，但参数量小10倍以上71% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1030798API
curl https://kongchang.com/api/v1/knowledge/claims/1030798MCP
get_claim(id=1030798)