Expired50% confidenceBenchmarkExact time
Using NVIDIA B300 chips in a 1-megawatt data center can achieve a total throughput of approximately 450,000 tokens/s, which at 340 tokens/s per user amounts to approximately 1,324 concurrent active users.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
7/21/2026
First Seen
Valid until: 10/19/2026(expired)
Sources
Related Claims
UnverifiedThe NVIDIA A100 has a memory bandwidth of 2 TB/s; loading a 280 GB FP32 model takes approximately 140 ms, meaning weight loading alone limits generation to a maximum of approximately 7 tokens per second.72% similarUnverifiedThe HBM memory bandwidth of the NVIDIA H100 is approximately 3.35 TB/s, and its utilization rate in autoregressive generation tasks is often below 10%.70% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/580876API
curl https://kongchang.com/api/v1/knowledge/claims/580876MCP
get_claim(id=580876)