Create a recorded presentation on a scientific discipline that uses high performance computing. The presentation should be 8-12 minutes long and should target your instructor and fellow class members. If you are involved in research that involves high performance computing, we prefer that your presentation be on that research. Otherwise, pick an interesting scientific topic that involves high performance computing–for example, you could present on engine design, black hole research, or genome sequencing.

You’ll spend about half your time talking about the topic of interest and half of your time explaining its computational needs and how high performance computing enables it. You should not assume that we know much about the domain you are presenting on.

You needn’t create a multimedia masterpiece. We’re looking for a basic slide presentation, much like one that you might deliver in an in-person class. If you find yourself using stock footage and background music, you’ve probably gone too far.

Your slides should be legible and your audio clear, but you won’t be graded on audio or visual quality unless it impairs understanding–recording a Zoom meeting while sharing your screen and going through the slides will probably be good enough. There are rooms in the library that you can use for recording if you’d like.

Submission

Upload your recorded video somewhere public and post the link to it in the #presentations slack channel. You can use Box (which is free for students) for this–just make sure to create a shared link that anyone can see if you use it. You can of course use whichever video hosting platform you’d like–if you’re already comfortable with one, feel free to use it as long as the video is publicly accessible.

Grading

We haven’t needed to make a rubric so far since almost without exception presentations have been relevant, engaging, and have followed the instructions–please don’t be the one that changes that.

You will be awarded 100% as long as you:

  • Put in reasonable effort
  • Follow the instructions
  • Provide meaningful commentary on both the domain specifics and the computational profile, roughly half the time spent on each
  • Turn it in on time