per-GPU HBM bandwidth, HBM3E (B200/B300 class) → HBM4 (Rubin, 8 stacks) — ~2.8x
~8 → ~22 TB/sobserved
| Value kind | observed — Reported measurements, counts, and specifications keep the precision and scope stated by their source; an exact specification is not treated as a range. |
|---|---|
| As of | 2026 |
| Source | NVIDIA; Tom's Hardware; TrendForce · Rubin-versus-Blackwell comparison chart; HBM bandwidth rises from 8 TB/s to as much as 22 TB/s |
| Review | checking…review by 2026-11-01 · standard cadence |
| Recorded changes | last 2026-06-29 |
| Claim id | per-gpu-hbm-bandwidth-hbm3e-b200-b300-class |
Where the guide uses it
← Full numbers register — every date-stamped figure in the guide, with revision history.