Cloud Computing Resources for Education
In this article, we explore the potential of cloud computing in eduction, including GPUs for students, academic labs, research groups, student-run orgs, and more.
Created on December 23|Last edited on February 28
Comment
Are you a student, TA/GSI, professor, or other person in academia who works with machine and/or deep learning models?
Is your lab, research group, or student-run club in need of cloud compute credits or substantial volumes of GPU credits?
To help your lab in performing the highest quality research we've compiled several "for education" cloud compute programs that you can sign up for: they're free for academics, Weights & Biases integrates with all the major cloud providers, and when used in tandem with our experiment tracking ecosystem you'll be able to accelerate your machine and deep learning like never before!
Google Cloud for Researchers from Google Cloud Platform (GCP)
Google Cloud for Researchers is a free cloud computing program that provides compute time, GPU hours, and other cloud computing resources to academics.
With this program, Google provides a powerful set of tools to enable researchers across academic domains to quickly and easily access their cloud computing needs. From molecular biology to reinforcement learning powering the robots of tomorrow, Google Cloud for Researchers helps today's students and educators power the innovation of tomorrow.
The program is open to all academics who have an active Google account and is completely free of charge. All that is needed to apply is to provide a few details about your research project and a bit of information about yourself. Google provides this program to help academics accelerate their research, so they can focus on the research itself and not be slowed down by time-consuming infrastructure or costs.
- Take free Google Cloud Skills Boost classes to acquire new skills in the GCP ecosystem
- Connect with a community of peers doing cutting-edge research
Additional hands-on labs for students to learn more cloud computing tools and infrastructure set-up can be found here: https://cloud.google.com/edu/students
Amazon Cloud Credits for Research from Amazon Web Services (AWS)
Similar to the programs offered by Google there are programs for academic research, education, student-run research, and more from Amazon that provide you with free access to:
- General cloud computing services (AWS)
- GPU compute credits
- Access to special types of hardware that may not be otherwise accessible to students, labs, etc.: Field Programmable Gate Arrays
Student awards are capped at $5,000 whereas faculty and staff awards are uncapped. Be sure to plan in advance as the review cycle takes 90 to 120 days. You can learn the answers to some frequently asked questions about the Cloud Credits for Research program here.
Like Google, Amazon requires you to submit a brief application to be considered for the AWS Cloud Credits for Research program:
The AWS Cloud Credits for Research program (formerly AWS Research Grants) supports researchers who seek to • Build cloud-hosted publicly available science-as-a-service applications, software, or tools to facilitate their future research and the research of their community. • Perform proof of concept or benchmark tests evaluating the efficacy of moving research workloads or open data sets to the cloud. • Train a broader community on the usage of cloud for research workloads via workshops or tutorials.https://aws.amazon.com/education/F1-instances-for-educators/
If your organization is a non-profit organization as some student-run clubs are Amazon does have non-profit aid programs; learn more here: https://aws.amazon.com/government-education/nonprofits/
Jarvis Labs' Special Pricing for Students, Social Causes, and Open-Source Creators
Jarvis Labs, the GPU provider that powers Kaggle and many other organizations in the deep learning space, offers special pricing if you're a student, working for a social cause, or are building open source tools.
Follow the instructions on the linked webpage, making sure to create an account on the JarvisLabs website before you submit an application for lower-cost GPU compute and storage.
Note that you cannot use spot instances (also called pre-emptible instances) as part of the special pricing plan; spot instances are about 50% of the cost of non-spot instances, so make sure to budget accordingly.
Paid Resources - Cloud Cost Comparison
If your lab or research group does already have funding that can be used to purchase cloud GPU credits here is a comprehensive GPU pricing list from all the major providers – AWS, GCP, JarvisLabs, Lambda, PaperSpace and more – from our friends at 🥞 Full Stack Deep Learning 🥞 :
Weights & Biases: In the Cloud, On Prem
Dedicated Cloud Deployments
After you've received compute credits from one of the programs above you'll be able to make use of Weights & Biases on all major cloud compute platforms. Check out our documentation to learn more about our Dedicated Cloud Deployment option!
Simply have your Terraform ecosystem hook into your data storage bucket and you'll be up and running with Weights & Biases in a few minutes!
On-Prem Deployments
For more information on how to have Weights & Biases run on hardware that you control please see our On-Prem / Bare Metal documentation page.
Private Cloud Deployments
You may want to leverage the flexibility of cloud-based deployments, avoiding the headaches that commonly occur with maintaining your infrastructure in an on-prem deployment situation, but you still require additional security policies that a dedicated cloud system does not offer. For you we offer a Private Cloud Deployment ecosystem which you can learn about here on our Private Cloud page.
Bonus: Github's Student Developer Pack
Once you have a cloud compute credits from the providers above you may want to obtain a set of developer tools from Github: https://education.github.com/pack
Software or cloud credits included in the Github Student Developer Pack include:
- Databases: MongoDB, PopSQL, SQLGate
- Deployment, Testing, CI/CD: Heroku, Sentry, DigitalOcean, Travis CI, LambdaTest, Codecov
- a wide range of interview prep tools: DataCamp, AlgoExpert.io, InterviewCake
Weights & Biases: Your Automated Logging Companion for Reproducible Research
How W&B helps students, researchers, teachers, and academics of all stripes log research for easy collaboration, sharing, and reproducibility
How Many Discoveries Were Lost Because They Weren't Written Down?
With Reports you can save your ingenious ideas and insights alongside a snapshot of your work and then share them with your team. Or just keep them all for when you need them.
Collaborative Research and Publication-Ready Graphics with W&B
Real-time collaboration and publication-ready graphics with a few mouse clicks or lines of code
DALL-E Mini Explained
Generate images from a text prompt in this interactive report: DALL-E Mini Explained, a reproduction of OpenAI DALL·E
Add a comment
Iterate on AI agents and models faster. Try Weights & Biases today.