# Create a GPU Instance

## Procedures

1. Log in to [**zenConsole**](https://console.zenlayer.com/), go to **Products** > **Virtual Machine** > **GPU & AI** > **GPU Cloud**, and click **Create GPU Instance**.
2. **Select a location**.\
   Choose the closest region for optimal performance.
3. **Choose your use case and image**.\
   Choose an available pre-built environment to quickly set up the instance with essential tools.
   * **AI/ML Inference**\
     Ollama image preinstalled with: `Llama 3.1:8b`, `NvidiaDriver 550.90.07`, `CUDA 12.6`, `cuDNN 9.3.0`, `Python 3.12.3`, `Ollama 0.3.9`, `JupyterLab 4.2.5`. Used for deploying models and running inference tasks.
   * **Common Use**\
     Ubuntu Server 24.04 LTS image preinstalled with: `NvidiaDriver 550.90.07`, `CUDA 12.6`, `cuDNN 9.3.0`. Used for common computing needs.
4. **Select a GPU model**.\
   Opt for the **NVIDIA GeForce RTX 4090** with 24 GiB of vRAM, ensuring high-performance computation with **82+ TFLOPS** of FP32 compute power, perfect for demanding AI tasks.
5. **Configure storage**.

   Set your storage requirements, basic and standard NVMe SSDs for fast I/O operations.
6. **Manage instance access**.
   * Add an **SSH key pair** to securely access your instance.
   * Use the default user **"ubuntu"** or create a custom password for login.
7. **Choose quantity**.\
   Set the number of instances you want to create and give the instance a **resource name**.
8. **More settings**.
   * Select your **OS time zone**.
   * Assign your resources to your desired **resource group**.
9. **Confirm order**.\
   Check the order summary and confirm your order.

## Key Considerations

1. **Location and Latency**: Choosing the right region is important for reducing latency. Select a region closest to your user base or data source.
2. **Storage Allocation**: Ensure you allocate sufficient NVMe SSD storage based on your workload requirements. If working with large datasets, you might need to expand storage.
3. **Security**:
   * Recommended to use SSH key pairs for secure access.
   * Regularly update your software and patches to maintain security.


---

# Agent Instructions: Querying This Documentation

If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.

Perform an HTTP GET request on the current page URL with the `ask` query parameter:

```
GET https://docs.console.zenlayer.com/test/wen-dang-zhong-xin-ce-shi-fen-zhi/gpu-cloud/virtual-machines/create-a-gpu-instance.md?ask=<question>
```

The question should be specific, self-contained, and written in natural language.
The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.