> ## Documentation Index
> Fetch the complete documentation index at: https://www.thundercompute.com/docs/llms.txt
> Use this file to discover all available pages before exploring further.

# Development vs Production

> Differentiate development and production environments on Thunder Compute. Select hardware and configurations optimized for your project scale.

Thunder Compute offers two modes for running instances.

| Feature       | Development                                    | Production                                                   |
| ------------- | ---------------------------------------------- | ------------------------------------------------------------ |
| Use case      | R\&D, experimentation, short-lived development | Long-running inference, batch training, production workloads |
| Cost          | Lower                                          | Higher                                                       |
| Compatibility | Most ML workloads                              | Full                                                         |

## Development Mode

<Note>
  Development mode is optimized for R\&D, experimentation, and short-lived development workloads. Use production mode for long-running inference services or batch training jobs.
</Note>

Development mode applies CUDA-level optimizations to maximize GPU utilization, significantly reducing costs for AI/ML development workflows.

### Supported Workloads

* **Research & Development**
* **Fine-tuning**
* **Training**
* **Small-scale inference**
* **Example software**: PyTorch (fully supported; downgrading from the pre-installed version may cause issues), TensorFlow, JAX, Jupyter Notebooks, ComfyUI, Ollama, VLLM, Unsloth

### Unsupported Workloads

* **Long-running production inference**: persistent inference servers, always-on APIs, or latency-sensitive serving
* **Batch training**: unattended production training jobs, scheduled training pipelines, or other long-running training workloads
* **Graphics workloads**: OpenGL, Vulkan, FFMPEG
* **Hardware-specific profiling tools**: tools that require direct hardware metrics or low-level device access

<Tip>
  If you encounter issues with an unsupported workload, switch to production mode with [modify](/vscode/operations/modifying-instances) for full compatibility.
</Tip>

## Production Mode

Production mode provisions a standard virtual machine with full CUDA compatibility and predictable performance.

### When to Choose Production

* Long-running training jobs
* Multi-GPU workloads (up to 8 GPUs)
* Graphics workloads (OpenGL, Vulkan, FFMPEG)
* Custom CUDA kernels
* Hardware profiling

## Switching Between Modes

[Modify existing instances](/vscode/operations/modifying-instances) to switch between development and production mode. This also lets you change GPU type, vCPUs, and RAM. Storage can be expanded but not reduced.

## Learn More

* [Technical Specifications](/technical-specs): Hardware, networking, and storage details