Home | Hyak

Your Docusaurus site did not load properly.

A very common reason is a wrong site baseUrl configuration.

Current configured baseUrl = / (default value)

We suggest trying baseUrl =

Machine Learning
We have tools and training to get you started with deep learning.
Molecular Dynamics
MD is one of the biggest use cases for HPC.
Computational Social Sciences
HPC is increasingly used in the social sciences.
Research Containers
We provide support for computational reproducibility and portability.
Machine Learning
We have tools and training to get you started with deep learning.
Molecular Dynamics
MD is one of the biggest use cases for HPC.

A user interface for everyone

We support a wide range of user interfaces, whether you're a seasoned command-line interface (CLI) user who only needs to know the hostname to SSH to, a novice who likes the comfort of a graphical user interface (GUI) desktop, or maybe you just prefer the convenience of a modern web-browser based experience.

We serve a web-based Visual Studio Code instance for developers (also emacs or vim if that's your thing). We have Jupyter Notebooks and RStudio server (both browser based) for data scientists interested in using bigger infrastructure. Even an Xfce GUI for a full on Linux desktop.

We speak your language

Any programming language that runs on an x86-64 processor can be supported on Hyak. The predominant ones are C/C++ as a part of widely deployed scientific applications or Python and R for data science and analytics. If you're a C/C++ or Python user, we have a license for Intel Parallel Studio with optimized compilers (e.g. icc) to speed up those codebases.

Other languages such as Go, Perl, Julia, and countless others are used every day on Hyak and you're welcome to bring your own to the community and make the platform yours.

Research App Store

Not a developer? Not a problem! There is a rich scientific app store on Hyak. There are are a number of Hyak maintained modules like MATLAB. Community maintained compiled software modules include things like Gaussian, star-ccm, and bowtie. We're also seeing increasing use of machine learning frameworks like TensorFlow and PyTorch and we encourage each user to maintain their own versioned library using Anaconda.

Interested in computational best practices? Don't want to be locked into a platform? The best option is to use containers for your code. We specifically support Apptainer which is usually interoperable with the more ubiquitous Docker. By developing your code locally in a Apptainer container, it becomes easily portable to Hyak or beyond to national supercomputing centers and the cloud. Your research goes wherever you go.