Unverified85% confidenceFactTime unknown
8卡DGX B200系统的GPU间总带宽达到14.4TB/s
1
Sources
85%
Confidence
Long-term
Relevance
5/31/2026
First Seen
Sources
Related Entities
Related Claims
VerifiedThe 3080 Ti has a memory bandwidth of 912 GB/s, far exceeding the 16–32 GB/s PCIe bus bandwidth between CPU and GPU.80% similarUnverified单台DGX H100服务器内8块GPU的双向带宽可达900GB/s77% similarVerifiedThe HBM bandwidth of the A100 GPU is approximately 2 TB/s, while the on-chip SRAM bandwidth can reach 19 TB/s, a difference of nearly 10x.75% similarVerifiedOn traditional x86 platforms, the bandwidth ceiling of the PCIe bus connecting a discrete GPU to the motherboard memory is approximately 64GB/s.74% similarUnverifiedThe 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.74% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/4499API
curl https://kongchang.com/api/v1/knowledge/claims/4499MCP
get_claim(id=4499)