Verified90% confidenceFactExact time
The HBM3 memory used by NVIDIA's H100 has a bandwidth of up to 3.35 TB/s.
9
Sources
90%
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
VerifiedThe GB300 integrates HBM3e high-bandwidth memory, with a single-card capacity of up to 288GB and memory bandwidth exceeding 10TB/s85% similarVerifiedThe H100 SXM5's HBM3 bandwidth is approximately 3.35 TB/s, and its FP16 peak compute is approximately 989 TFLOPS.84% similarUnverifiedNVIDIA's H100 is based on the Hopper architecture, uses HBM3 memory with a bandwidth of up to 3.35 TB/s, and delivers single-card FP16 compute of nearly 2,000 TFLOPS.84% similarUnverifiedNVIDIA's GB300 is the flagship of the Blackwell Ultra architecture, with a single chip featuring 288 GB of HBM3e memory capacity and total memory bandwidth exceeding 16 TB/s, an improvement of nearly 5× over the H100's 3.35 TB/s.79% similarUnverifiedNVIDIA H100搭载的HBM3提供约3.35TB/s内存带宽，而DDR6预计单通道最高带宽约256GB/s79% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/489298API
curl https://kongchang.com/api/v1/knowledge/claims/489298MCP
get_claim(id=489298)