Immediate access to NVIDIA H200 Tensor Core GPUs
From a single GPU to thousands. Choose on-demand, pay-as-you-go options or cost-effective reserved plans.
The most powerful GPU for your AI and HPC workloads
Prices for NVIDIA H200 Tensor Core GPU
Contact our sales team to get the best pricing and a free proof-of-concept stage for your AI project.
$2.30 GPU/h
As low as $2.30 per 1 GPU/hour, the NVIDIA H200 with Intel Sapphire Rapids is available now.
$2.59 GPU/h
Price for 1 GPU/hour with a minimum of 1-month reserve commitment to a minimum of 8 GPUs.
NVIDIA H200 vs NVIDIA H100
NVIDIA H200 vs NVIDIA H100
According to an NVIDIA research, NVIDIA H200 shows next-level performance compared to NVIDIA H100 for popular LLMs: Llama2 13B, Llama2 70B and GPT-3 175B.
NVIDIA H200 SXM offers significant enhancements over the NVIDIA H100 SXM, delivering up to 45% better performance in generative AI and HPC tasks.
![](https://nebius.directus.app/assets/dd802957-37ce-4e05-9b1e-c7a8261e13bf/pages-h200-scheme-2.png?cache-buster=2024-12-10T10:03:38.950Z)
NVIDIA GPU clusters with InfiniBand
NVIDIA GPU clusters with InfiniBand
We provide NVIDIA GPU clusters with InfiniBand interconnect.
InfiniBand is a technology that facilitates seamless connection of servers equipped with NVIDIA GPUs within a cluster. Leverage multi-node interconnection with up to 3.2 Tbit/s.
![](https://nebius.directus.app/assets/90e5d2ae-d85e-4f2a-a7f9-9aa98ab61bf1/pages-supercomputer-hardware.jpg?cache-buster=2024-12-10T14:44:56.201Z)
How the committed reserve works
Place an order now to lock in the price for your future GPU needs.
Our sales team will contact you soon to confirm the dates and provide the contract, which must be signed at least one month before your reserve starts. You can adjust the reserve dates anytime up to a month before the start.