Below is a summary of the kinds of technology resources and services that SEAS Computing provides in support of courses at the John A. Paulson School of Engineering and Applied Sciences.
SEAS can provide consultation and development of computational content for courses geared towards engineering sciences. This includes creating and presenting in-class demos, tutorials, and computational lab sections. The content will be handed over to the course after development and SEAS Computing will work with teaching staff on updates and support.
1. AWS EC2 Single Server (for advanced users only)
SEAS Computing can create individual EC2 instances, hosted within the SEAS AWS infrastructure, and hand them off to course staff. This is meant for more advanced users. The ongoing cost of these EC2 instances will need to be covered by the course.
If multiple EC2 instances are required, SEAS Computing can create a template (AMI) to ensure that the machines are built with a common configuration. We also deploy scripts for managing resources, such as alerting to maximum CPU usage and shutting down EC2 instances when they are not used.
2. AWS Educate
SEAS Computing can provide consultation and documentation to teaching staff to help them obtain credit for both teaching staff and students through the AWS Educate program. Instructions for how to apply for AWS Educate credit can be found on the SEAS wiki.
3. Course Organization Apps
SEAS provides a set of Canvas LTI applications for aiding in the organization of courses. These tools are geared towards computer science courses, but are available to all SEAS courses. Tools are self-service with support from Jonathan Seitz. The tools integrate with Canvas and the Speedgrader tool. For more details visit inclass.seas.harvard.edu.
4. Git Code Repository
Students are able to claim an account on Harvard's institutional GitHub implementation, https://code.harvard.edu. Authentication is provided via HarvardKey. SEAS's own code repository (code.seas.havard.edu) is currently being phased-out and will be retired during the 2019/20 Spring semester.
5. High performance computing (HPC) with multiple GPUs
Some classes, e.g. CS109b, need HPC and GPUs. SEAS Computing can provide guidance on how to get access to these resources. We can help with:
Cloud GPU resources such as EC2 instance type GPU. Amazon provides a wide range of GPU machine specifications.
Physical GPU resources such as those on the Odyssey cluster.
In the case of Cloud GPU resources, SEAS Computing can work with teaching staff to define course needs and build and support the relevant AWS infrastructure for a course, provided that the cost is covered by the course
6. RC Class Accounts
RC is currently accommodating classes as and when needed. The process requires for the class to send a ticket to RC listing the first and last names, and email addresses of the students and total. The accounts are destroyed and the data is erased at the end of the semester.
Jupyter Notebook is a tool that extends a physical machine's access to interactive computing by providing access to a ready-to-go web-based environment with the necessary software tools to enable students to develop, document, and execute code, as well as communicate the output for a course. JupyterHub allows for this consistent environment to be made accessible to all students in a course.
JupyterHub as a service is now managed by the FAS Academic Technology Group.
The application is setup by FAS ATG and integrated with Canvas, so there is no setup required by the student. Students can concentrate on course content from the first day of classes. FAS ATG can provide consultation on requirements and implementation of the JupyterHub infrastructure.
To request a JupyterHub implementation for your course, please fill out the following form: http://bit.ly/fas-jupyter-request
8. Windows Workspaces - Virtual Desktop Images (VDI)
SEAS Computing provides cloud-based virtual Windows desktops hosted in Amazon's AWS.
Students can use them for a semester, and can run applications such as Comsol, Abaqus, and SolidWorks remotely from their laptops. SEAS Computing can provide consultation on setup and pricing.
For courses where students may require access to a physical machine, SEAS Computing is able to consult on availability of such resources. SEAS Computing can also facilitate introduction to the Active Learning Labs where students have access to physical machines.