Here are a few ways to get FREE access to Machine Learning & GPU computing resources

• AWS Free Tier account provides 2 months free use of many Amazon services, such as SageMaker, to learn or explore machine learning.
• AWS SageMaker Studio Lab provides a free notebook environment for individuals that want to learn and experiment with machine learning code and models. Key benefits are:
  • Completely free, you only need a valid email - no credit card or AWS account required
  • No Set Up required - enabling you to focus on the data science lesson, not the configuration headaches
  • Based on the open source community Project Jupyter
  • Access to both CPU (12 hours per user session) and GPU (4 hours per user session)
  • 15 GBs of persistent storage - enabling you and your students to save work in between user sessions
  • Integration to Github
  • For instructors who want to use SageMaker Studio Lab in your classrooms, AWS can issue additional reference codes to enable students to also get access immediately. For more information, contact Luke Chirhart.
• AWS Credits:
  • AWS Activate provides credits for startups. If you are working toward a solution for possible commercialization you can apply for $1000 of credits. 
    
  • AWS credits for Proof-of-Concept (POC) projects. You can contact Luke Chirhart, chirhart@amazon.com, our Account Manager at AWS for some AWS credits to experiment with AWS services toward a research project. Funding is typically for three months to validate the project.
  • AWS Cloud Credit for Research provides larger amounts of credits for research projects after the POC phase. Submit brief, single-page proposals to get short term credits for AWS usage. No limits on faculty awards. Student awards are limited to $5000. For more information, see AWS Cloud Credit for Research FAQs.
• AWS Training
  • AWS Educate is a free, self-paced environment to help you learn cloud technology. See the AWS Ramp-Up Guide: Machine Learning for a list of courses.
  • AWS Academy provides free access to AWS but only through a series of courses provided by AWS that are part of an existing course. Instructors wanting to incorporating AWS Academy materials in their courses can find more information here.
• NIH & NSF AWS Research Initiatives (ARI) – Submit research project proposals to get funding from NSF or NIH to use AWS compute resources.
• NSF FutureCloud program has funding for their Chameleon HPC environment: https://chameleoncloud.org/learn/frequently-asked-questions/.
• CloudLab is available for research and education use: https://cloudlab.us/.
• NIH STRIDES offers $500 credits for short-term projects on AWS, Azure, or GCP.
• NCSA offers computing services through the Access program: https://www.ncsa.illinois.edu/research/project-highlights/access/
• NCSA offers computing services through the Illinois Computes program: https://www.ncsa.illinois.edu/about/illinois-computes/
• Open Science Grid offers GPU use through their OSPool of distributed High Throughput Computing (dHTC) services. https://portal.osg-htc.org/documentation/htc_workloads/specific_resource/gpu-jobs/
• Code Love blog lists several free access solutions for GPUs: https://code-love.com/2020/08/08/where-to-get-free-gpu-cloud-hours-for-machine-learning/.
• Google Colab provides access free of charge to computing resources: https://colab.research.google.com/notebooks/intro.ipynb
• After using Google Colab you can move to Google Cloud GPUs with $300 of credit: https://cloud.google.com/gpu
• Google TPU Research Cloud provides grants for ML research projects: https://sites.research.google/trc/about
• Kaggle.com provides 30 hours free usage per week: https://www.kaggle.com/docs/efficient-gpu-usage
• Microsoft Azure provides $200 of credit: https://docs.microsoft.com/en-us/azure/virtual-machines/sizes-gpu
• Gradient by Paperspace offers free usage of GPUs: https://www.paperspace.com/gradient/free-gpu

Questions: Send an email to cloud@illinois.edu or aws-support@illinois.edu