H100-NVL-94GB vs H100-PCIE-80GB
H100-PCIE-80GB is the cheaper of the two: $1.66/hr at Shadeform, against $3/hr for H100-NVL-94GB at Vast.ai — 1.8× the price.
Cheapest on each
Aligned facts
Same sheet, same order, both cards — including the gaps.
| Attribute | H100-NVL-94GB | H100-PCIE-80GB |
|---|---|---|
| VRAM | 94 GB▲ | 80 GB |
| Bandwidth | 3.9 TB/s▲ | 2 TB/s |
| FP16 | 835 TFLOPS▲ | 756 TFLOPS |
| FP8 | 1670 TFLOPS▲ | 1513 TFLOPS |
| TDP | 400 W▲ | 350 W |
| NVLink | NVLink 4 | PCIe 5 (bridge)▲ |
| Architecture | Hopper | Hopper |
| Released | 2023▲ | 2022 |
A blank is not a zero. Where one card states a figure and the other states none, neither wins the row — the second number is unknown, and ranking an unknown as nothing is how a comparison starts to lie.
Every seller, both cards
One row per seller, cheapest tier, on-demand list prices only. Sellers of both cards lead, cheapest combined rate first.
| Seller | H100-NVL-94GB | H100-PCIE-80GB |
|---|---|---|
| Vast.ai | $3.002026-10-27 | $2.542026-10-27 |
| RunPod | $3.192026-10-27 | $2.892026-10-27 |
| Massed Compute | $3.112026-10-26 | ○ no answer |
| Shadeform | ○ no answer | $1.662026-10-26 |
| Beam | ○ no answer | $1.832026-10-27 |
| Cyfuture AI | ○ no answer | $2.992026-10-27 |
| Lambda | ○ no answer | $3.292026-10-26 |
| Oblivus | ○ no answer | $3.502026-10-26 |
What this page will not tell you
On-demand list rates only · spot and committed terms priced separately and not mixed in here · no performance-per-dollar figure appears on this page, because it needs a throughput benchmark we do not hold and dividing a price by a vendor’s own TFLOPS claim produces a number that looks authoritative and means nothing