Datacenter flagship · Hopper · launched 2022 · prices checked 2026-09-03
Rent NVIDIA H100 SXM — 80 GB, $1.587/hr on-demand
- VRAM 80 GBHBM3
- FP16 tensor 990TFLOPS
- PowerScore 697RTX 3090 = 100
- Configs 1–8×NVLink
- Online now 69 7 regions
Billed per second, price locked at deploy. Storage $0.08/GB/mo · bandwidth $0.01/GB — the whole fee schedule. Next weekly market re-check: 2026-09-10.
Datacenter flagship · Hopper architecture
The H100 SXM is the reference training GPU of its generation: 80 GB HBM3 at 3.35 TB/s, 700 W and full 900 GB/s NVLink on 8-GPU HGX baseboards. It is the deepest pool on the sheet, which keeps interruptible slots available, and the price-performance default for pre-training, RLHF and full fine-tunes.
This is training-grade silicon: HBM3 memory feeding tensor cores at multi-TB/s, NVLink for scaling past one card, and the reliability profile of Tier-III datacenter hosts. Teams rent it for pre-training, long fine-tunes and high-throughput inference where batch size is money.
NVIDIA H100 SXM specs: VRAM, TFLOPS, bandwidth
| GPU model | NVIDIA H100 SXM | Architecture | Hopper (2022) |
|---|---|---|---|
| VRAM | 80 GB HBM3 | Memory bandwidth | 3,350 GB/s |
| FP16 tensor perf. | 990 TFLOPS | FP32 perf. | 67.0 TFLOPS |
| CUDA cores | 16,896 | TDP | 700 W |
| PowerScore (RTX 3090 = 100) | 697 | PCIe generation | Gen 5.0 |
| Multi-GPU | 1× – 8× · NVLink | Max instance storage | 16,000 GB NVMe |
| Network up to | 10,000 Mbps | CUDA | 12.4 – 13.0 |
Bandwidth, CUDA cores, TDP and FP32 are public NVIDIA figures; FP16 tensor is the dense (non-sparsity) number. Machine-level values come from live inventory.
H100 SXM price per hour: on-demand, interruptible, reserved
One public rule sets every price on this page: the marketplace median for the H100 SXM ($2.27/hr, snapshot 2026-09-03) × 0.70, rounded down — so on-demand is $1.587, 30% below market. Interruptible halves it; a 3-month reservation takes another 35% off.
| Mode | Per GPU-hour | Per day (24 h) | Per month (730 h) | What you get |
|---|---|---|---|---|
| On-demand | $1.587 | $38.09 | $1,159 | Guaranteed capacity, price locked at deploy, stop anytime |
| Interruptible | $0.793 | $19.03 | $579 | Flat −50%; may pause under capacity pressure, disk kept, auto-requeue |
| Reserved (3 months) | $1.031 | $24.74 | $753 | −35% on on-demand, rate locked for the term, capacity held |
Per GPU: an 8× machine costs exactly 8× — no multi-GPU premium. Estimate a full month with storage and bandwidth in the GPU cost calculator.
What you can run on a H100 SXM (80 GB VRAM)
With 80 GB of HBM3, a single card holds a ~32B-parameter LLM in FP16 or up to ~123B parameters quantized to 4-bit, with room for KV-cache at practical context lengths. Scale to 8× GPUs on one machine with NVLink for bigger models or bigger batches — the per-GPU price stays $1.587.
- Recommended for llm training — H100 SXM from 30%+ under market
- Recommended for scientific computing — Up to 4.8 TB/s per GPU
- One-click template: vLLM on a H100 SXM
- One-click template: PyTorch NGC on a H100 SXM
- One-click template: Axolotl — Fine Tuning on a H100 SXM
- Sizing help: LLM VRAM requirements guide
H100 SXM availability by region
69 × H100 SXM across 12 machines, live from inventory:
Amsterdam
Stockholm
Mumbai
Montréal
London
Chicago, IL
Singapore
H100 SXM vs alternatives: price per TFLOP
| GPU | VRAM | FP16 | On-demand | $ / TFLOP-hr |
|---|---|---|---|---|
| H100 SXM this card | 80 GB | 990 | $1.587 | $1.60‰ |
| A100 SXM4 | 80 GB | 312 | $0.583 | $1.87‰ |
| H200 | 141 GB | 990 | $3.058 | $3.09‰ |
| B200 | 192 GB | 2,250 | $4.204 | $1.87‰ |
| B300 | 288 GB | 2,800 | $7.875 | $2.81‰ |
‰ = dollars per 1,000 TFLOP-hours of FP16 — a rough value-for-compute yardstick across cards.
- H100 SXM vs A100 SXM4 — specs, price per hour, which to rent
- H100 SXM vs H200 — specs, price per hour, which to rent
- H100 SXM vs B200 — specs, price per hour, which to rent
- H100 SXM vs H100 PCIE — specs, price per hour, which to rent
- H100 NVL vs H100 SXM — specs, price per hour, which to rent
- All GPU comparisons
Renting a H100 SXM: frequently asked questions
How much does it cost to rent an NVIDIA H100 SXM per hour?
$1.587 per GPU-hour on-demand — a fixed price set at least 30% below the current market median of $2.27. Interruptible capacity costs $0.793/hr and a 3-month reservation $1.031/hr. Around $1,159/month if you keep one running non-stop, billed per second.
What can a H100 SXM with 80 GB VRAM run?
In LLM terms, roughly a 32B-parameter model in FP16 or up to ~123B parameters 4-bit quantized on a single card, with context headroom. Multi-GPU instances (up to 8× on current inventory) multiply that; diffusion and rendering workloads fit comfortably at this VRAM class.
Is the H100 SXM available to rent right now?
Yes — 69 GPUs across 12 machines in 7 regions are listed as we render this page. Configurations go from 1× to 8× with NVLink on multi-GPU chassis. Deploy from the console and it is running in about 30 seconds.
How do I deploy a H100 SXM?
Create an account (email + password, no card, no KYC), top up in crypto, open the console, filter by H100 SXM, pick a machine and a template such as PyTorch, vLLM or ComfyUI. The same deploy is one command with the CLI: powergpu launch --gpu h100-sxm --template pytorch.
Why is the H100 SXM cheaper here than on GPU marketplaces?
We price from the public marketplace median and fix our on-demand rate at least 30% below it, rounded down. The price is re-checked weekly (next check 2026-09-10) and published — no auctions, no per-host roulette, no bidding.