NVIDIA A16 vs NVIDIA H200 NVL
How the NVIDIA A16 and NVIDIA H200 NVL compare across memory, compute, power and form factor, with the stronger figure highlighted on each row.
| Specification |
AMPERE
NVIDIA A16
View Product →
|
HOPPER
NVIDIA H200 NVL
View Product →
|
|---|---|---|
| Pricing | ||
| Price | – | $41,400.00 |
| Overview | ||
| Architecture | Ampere | Hopper |
| Memory | ||
| Memory | 64 GB GDDR6 ECC (4× 16 GB) | 141 GB HBM3e |
| Memory Bandwidth | 4× 232 GB/s | 4.8 TB/s |
| Compute | ||
| CUDA Cores | 5,120 (4× 1,280) | – |
| Tensor Cores | 160 (4× 40) | – |
| RT Cores | 40 (4× 10) | – |
| FP32 | – | 67 TFLOPS |
| FP64 Tensor | – | 34 TFLOPS / 67 TFLOPS |
| FP8 Tensor | – | 3,958 TFLOPS (with sparsity) |
| Platform | ||
| Interface | PCIe Gen4 x16 | – |
| NVLink | – | NVIDIA NVLink (900 GB/s), PCIe Gen5 (128 GB/s) |
| TDP | 250W | Up to 600W (configurable) |
| Form Factor | Dual-slot, full-height, passive | PCIe dual-slot air cooled |
| Multi-Instance GPU (MIG) | – | Up to 7 MIGs @ 18 GB each |
| GPUs per Board | 4 | – |
| Decoders | – | 7 NVDEC, 7 JPEG |
Not sure which fits?
Talk to an Expert →
Talk it through with a specialist
Tell us your workload and we will recommend the right GPU, configuration, and sourcing options.