Skip to content

Welcome to JUNIQ

JUNIQ is an interactive, browser-based platform designed to provide seamless access to both supercomputing and cloud resources from the Jülich Supercomputing Centre (JSC). Whether you are working on large-scale simulations, machine learning models, or data analytics, JUNIQ gives you the flexibility to perform these tasks interactively through familiar Jupyter Notebooks.

For users with HPC accounts, you can harness the power of JSC's supercomputers like JUWELS, JEDI, JURECA and more. Additionally, users without HPC accounts can still benefit from the platform by utilizing JSC-Cloud resources.

Why Use JUNIQ?

By using JUNIQ, you gain access to:

1. Supercomputing and Cloud Power in Your Browser

JUNIQ connects you to world-class supercomputing systems such as JUWELS, JEDI, JURECA and more. But even if you don't have access to HPC resources, you can still use JSC-Cloud, offering scalable computing power for your workflows, including smaller data analysis tasks, web applications, and development environments.

2. Easy-to-Use Interface

The platform uses the popular Jupyter Notebook interface, making it easy for users of all skill levels to interact with powerful computational resources. Whether you're using HPC resources or the cloud, you can run code, develop models, and perform analyses all from your web browser.

3. Scalable Workflows

With JUNIQ, you can start with lightweight jobs on the JSC-Cloud or scale up to thousands of cores and GPUs on HPC systems for large-scale simulations and analysis.

4. Interactive Development

Develop and test code interactively with instant feedback, enabling fast iteration of data science workflows, simulations, and machine learning models. You can choose between the HPC infrastructure or the cloud based on your project's needs.

5. Use Custom Docker Images

JUNIQ supports custom Docker images, meaning you can define and use your own computing environment. This flexibility allows you to pre-install specific dependencies, configure your environment exactly as needed, and run reproducible computational workflows across multiple sessions.

6. GitHub-Based Environment with Binder

Through repo2docker integration, JUNIQ allows you to use the mybinder.org infrastructure to build Docker images directly from your GitHub repositories. This feature enables you to turn a GitHub repo into a fully functional environment without manually building Docker images, making your code and environment highly portable and shareable.

7. Access to Specialized Libraries and Tools

JUNIQ provides pre-configured environments with libraries optimized for both HPC and cloud use cases, including scientific computing and machine learning tools such as TensorFlow, PyTorch, and more.

Use Cases

JUNIQ is ideal for:

  • Large-Scale Simulations: Use HPC resources for complex simulations in fields such as physics, engineering, and climate science.
  • Machine Learning: Train machine learning models at scale using cloud or HPC resources.
  • Data Science and Analytics: Analyze and visualize large datasets interactively, leveraging cloud or HPC depending on your requirements.
  • Scientific Research: From bioinformatics to materials science, accelerate your research using cloud or supercomputing resources.
  • Workshops and Training: JUNIQ is a great platform for hands-on workshops and training sessions. Instructors can create custom environments, share notebooks, and guide participants through interactive coding exercises on both cloud and HPC infrastructure. This makes it ideal for educational and professional development settings.

Getting Started

To begin using JUNIQ, just follow these simple steps:

  1. Visit the website: Go to juniq.fz-juelich.de.
  2. Login: Use your JSC account credentials, or authenticate via Helmholtz AAI. More information here.
  3. Select a system: Choose from HPC systems like JUWELS, JEDI, JURECA, or the JSC-Cloud for cloud-based resources.
  4. Start and have fun: Launch your Jupyter environment and start working on your projects, whether they involve data science, machine learning, simulations, or training workshops.

Enjoy the power of supercomputing and cloud resources right from your browser!