The table places the range of both examples next to the GPU memory of each configuration. “Fits” means that the upper limit also fits, “Tight” means that only the lower limit fits; in tight cases, a load test decides.
For Example A, NG-L40S with 48 GB is the smallest configuration that fits. Example B requires NG-H100×2 with 160 GB of GPU memory in FP8; with weights in four bits, however, it fits on a single card.
GPU configurations against the two calculation methods| Configuration | Cards | Total GPU memory | Example A | Example B | Net monthly price |
|---|
| NG-RTX4000 | 1 × NVIDIA RTX PRO 4000 Blackwell SFF, 24 GB each | 24 GB | Not sufficient | Not sufficient | €214.00$248.24 |
|---|
| NG-L4 | 1 × NVIDIA L4, 24 GB each | 24 GB | Not sufficient | Not sufficient | €540.00$626.40 |
|---|
| NG-L40S | 1 × NVIDIA L40S, 48 GB each | 48 GB | Sufficient | Not sufficient | €1,008.00$1,169.28 |
|---|
| NG-H100 | 1 × NVIDIA H100 SXM, 80 GB each | 80 GB | Sufficient | Not sufficient | €1,940.00$2,250.40 |
|---|
| NG-RTX6000 | 1 × NVIDIA RTX PRO 6000 Blackwell Max-Q, 96 GB each | 96 GB | Sufficient | Not sufficient | €1,199.00$1,390.84 |
|---|
| NG-H100×2 | 2 × NVIDIA H100 SXM, 80 GB each | 160 GB | Sufficient | Sufficient | €3,880.00$4,500.80 |
|---|
| NG-H100×4 | 4 × NVIDIA H100 SXM, 80 GB each | 320 GB | Sufficient | Sufficient | €7,770.00$9,013.20 |
|---|
| NG-H200×8 | 8 × NVIDIA H200 SXM, 141 GB each | 1,128 GB | Sufficient | Sufficient | €30,660.00$35,565.60 |
|---|
The monthly prices are net catalogue prices for Frankfurt am Main. The RTX PRO 6000 is listed in the catalogue in several expansion stages that differ in memory and drives, not in the card; the table lists the cheapest. Each card belongs entirely to your node, without vGPU sharing, so the full GPU memory is available.
The catalogue lists multi-card configurations in the H100 and H200 class, with NVLink between the cards and up to 8 cards per node. A model is then distributed across the cards: each card holds its share of the weights and the KV cache, and each needs its own overhead. The total GPU memory is therefore an upper limit, not a guarantee.
Whether a model fits is determined by the size of the GPU memory; how fast it responds depends primarily on the memory bandwidth when generating the response, and the catalogue shows the memory type per card for this.
A few hours are enough for such a load test. Billing is the same as for all hourly billed lines: recorded to the second, billed per hour, capped monthly; without a term, you pay for the time used. A running card runs until you stop it; there are no interruptible instances at spot prices. Which cards start immediately and which are provided on request based on capacity can be found under the models on the GPU instances page.
The finder on the same page names the smallest card for the weights of your model and calculates the KV cache as a flat-rate overhead; you add long contexts and many concurrent requests using the formula above. If you only want to call one of the ready-made models of the AI platform, you do not need your own card: the AI platform bills per million tokens, with no base fee.