​
Supported GPUs

Compute Exchange currently supports the following NVIDIA GPUs for trading:

  • NVIDIA H100: High-performance GPU designed for deep learning training, inference, and HPC workloads. It is significantly more powerful than the A100 and optimized for large-scale AI models.

  • NVIDIA H200: An improved version of the H100 with better performance for AI training, inference, and HPC tasks. It is particularly well-suited for large-scale models like generative AI and LLMs, offering improved efficiency and throughput compared to the H100.

  • NVIDIA A100: Optimized for AI training, inference, and HPC applications but is now considered a previous-generation product compared to the H100 and H200.

In the coming months, we plan to introduce additional GPU models as they become available.

If you have specific compute requirements, reach out to us to discuss future offerings

​
Custom Compute Options

For specific compute configurations, users can:

  • Utilize filtering options in the marketplace to refine their searches.

  • Contact our support team to request specific compute types.