Search
Search the entire web effortlessly
maxresdefault 2025 04 08T201438.665
Run VS Code in the Browser with Google Cloud and TPU Power

The flexibility and power of Visual Studio Code (VS Code) make it a popular choice among developers. With the rise in cloud computing, hosting your development environment online can provide significant advantages. In this article, we will explore how to run VS Code in the browser, using Google Cloud Platform (GCP) and leveraging Tensor Processing Units (TPUs) for incredible computational power, particularly beneficial for machine learning tasks.

What is Coder?

Coder is a pioneering startup that has created a way to run VS Code in the cloud, allowing developers to access their integrated development environment (IDE) from anywhere—even on mobile devices. This capability is particularly advantageous for remote teams that require collaboration on projects in real time.

Benefits of Using Coder to Host VS Code

  • Accessibility: Access your coding environment from any device, anytime, without any setup.
  • Collaboration: Multiple developers can work on the same project within the same editor.
  • Resource Scalability: Attach an unlimited amount of computational resources to your IDE,
  • Customization: Retain all your favorite extensions and settings, just like on your local machine.

Setting Up Google Cloud Platform (GCP)

To begin, you will need a Google Cloud account. Follow these steps to create your own instance of VS Code in the cloud:

Step 1: Create a Virtual Machine (VM)

  1. Log into GCP and navigate to the Compute Engine section.
  2. Click on Create Instance and choose appropriate settings:
  • Opt for at least 2 CPU cores and 7.5 GB RAM to ensure comfortable performance.
  • Select a suitable operating system, preferably one with support for the required software.
  • Ensure to allow HTTP and HTTPS traffic for web access.
  1. Click Create and wait for your instance to be operational.

Step 2: Install Code Server

  1. After your VM is ready, access the command line via the SSH button next to your VM’s name.
  2. Use the following commands to install code server:
  • Download the binary for code server from the official GitHub repository.
  • Unzip the downloaded file and navigate into the directory.
  • Run the code server command with sudo privileges to start it on port 80.
  1. Take note of the password generated during this process, as it will be needed later for access.

Step 3: Access VS Code in the Browser

  1. Retrieve your VM’s external IP address from the GCP console.
  2. Append port 80 to this IP and navigate to it in your browser.
  3. You will receive a security warning; proceed by clicking on Advanced and then on Proceed to [your IP].
  4. Enter the password to access your VS Code environment, now fully operational in the cloud.

Enhancing Your Setup for Development

Front-End Development

For front-end development, you might want to set up additional services such as Angular or React.

  • Install Node.js using Node Version Manager (nvm).
  • Create your desired app and run it on the VM.
  • Firewall Settings: Remember to configure Google’s firewall settings to allow traffic through the appropriate ports (e.g., port 4200 for Angular) and verify that those rules are applied.

Machine Learning with TPUs

If you’re diving into machine learning, integrating TPUs into your workflow is a game-changer. TPUs are highly optimized for processing tensor operations, making them a better option for large-scale ML tasks than traditional GPU setups.

Step 4: Set Up a TPU

  1. Go to the TPU section within the Compute Engine.
  2. Create a TPU node alongside a new virtual machine that can interact with the TPU.
  3. Install TensorFlow on the VM, which is essential for machine learning tasks.

Performance Comparison: TPU vs CPU

To illustrate the difference between CPUs and TPUs:

  • CPU Model Training: Training a neural network on a basic VM may take around 90 minutes to process a certain dataset.
  • TPU Model Training: The same task on a TPU could potentially reduce the time to only 2.5 minutes.

Conclusion

Running VS Code in the browser using Google Cloud allows for remarkable flexibility, scalability, and collaboration for developers. The integration of TPUs also empowers ML enthusiasts to dramatically reduce the time needed for training models.

With tools like Coder and Google Cloud, development work is becoming more efficient and accessible, not only on traditional desktops but through robust cloud infrastructures.

Take advantage of these advanced tools to elevate your programming projects, explore machine learning capabilities, and collaborate seamlessly with your team.

    More posts