​
Use cases

Thunder Compute is optimized for AI/ML development workflows. That said, Thunder Compute has the full functionality of an EC2-style on-demand GPU cloud instance.

​
Technical specs

  • Egress/Ingress: 7Gbps
  • IP: dynamic
  • Region: U.S. Central (Iowa)
  • E or N series CPU instances in GCP

​
Officially supported libraries

The following libraries and tools are thoroughly tested:

  • PyTorch
  • PyTorch Lightning
  • Hugging Face
  • Notebooks
  • AI model serving tools like ComfyUI, Ollama, VLLM, and more
  • Tensorflow [experimental]
  • Jax [experimental]
  • Custom CUDA Kernels (message us on Discord if you are profiling and encounter errors)

Note: make sure you install the cuda-compatible version of these libraries. The cuda-compatible PyTorch binary and latest CUDA drivers are pre-installed on every Thunder Compute instance.

Do not attempt to reinstall CUDA. If it seems like you need an older CUDA driver, you almost always are better off upgrading your other dependencies (e.g., PyTorch)

You can create and manage Thunder Compute instances with the tnr command line interface from any major operating system. See our CLI reference for detailed usage instructions.

​
Limitations

We use a new kind of virtualization to drive down cost. To learn more about how this works, check out this blog post.

Currently, Thunder Compute lacks official support for graphics workloads such as OpenGL and Vulkan. If you’d like to run these, contact us.

If you encounter any issues or errors, please check our troubleshooting guide first.

​
Cryptocurrency mining

Mining, staking, or otherwise interacting with cryptocurrency is strictly prohibited on Thunder Compute. If cryptocurrency-related activity is detected, the associated account is immediately banned from Thunder Compute and any billing credit is revoked. The account is then billed for the full amount of usage.

To help you get started with Thunder Compute, we recommend checking out these guides: