Skip to content
HardwareLadder
AI GPU ladder

NVIDIAAI GPU · PCIe (NVL)

H100 NVL

NVIDIA H100 NVL is a Hopper data-center accelerator with 94 GB HBM3 and 3.9 TB/s of memory bandwidth.

Ladder rank
#9
AI index (H100 SXM = 100)
99.2
Series
Hopper

Specifications

Dense throughput, no sparsity
Architecture Hopper
Form factor PCIe (NVL)
Process TSMC 4N
FP16 / BF16 tensor (dense) 835.5 TFLOPS
FP8 tensor (dense) 1,670.5 TFLOPS
FP4 tensor (dense) —
Memory 94 GB HBM3
Memory bandwidth 3.9 TB/s
Power 350–400 W
Launch Mar 2023

AI index

Tensor × bandwidth

100 × √((835.5 ÷ 989.4) × (3.9 ÷ 3.35)) = 99.2

The AI index is the geometric mean of dense 16-bit tensor throughput and memory bandwidth, each relative to the H100 SXM. It is a spec-sheet comparison, not a measured benchmark.

Sources

2 references