It's pretty great to see students diving into serverless architecture head-on. We've been partnering and participating in a number of student hackathons with StdLib [1] and the amount students can achieve in less than 36 hours with our FaaS platform has knocked our socks off, kudos to AWS for kickstarting this transition. It's been wild to see the evolution of the "serverless" landscape just in the past 12 months, messaging has moved from cost-reduction (no over-provisioning) to time-saving and ease-of-use across the board. Seems that developers are finally starting to value the time it takes to learn and implement something and optimize for shipping value ASAP.
What's exciting to me is that this is just the beginning. These students are representative of the next generation of developers. So much innovation left to go; I believe we'll eventually look upon the "cloud" as merely an extension of our own computers, auto-scaling and responding to any workload seamlessly much like how nowadays we rarely give second thought to memory management, garbage collection, etc. when writing in JavaScript or Python.
Why is there no discussion on Docker or Kubernetes or Mesos or container technology in general? I hope the syllabus is revised to include discussion on these two topics, since all major cloud providers are investing much in container deployments—if only for a single day. Also, might be nice if they invited a speaker from Docker (I know a contact at Docker if class organizer needs one).
Interesting point, we did have a pretty good discussion about operations and VMs/containers in the guest lecture by Aaron Davidson from Databricks on 11/13 (slides/notes on the website).
Maybe half of this year's final projects focused on serverless/AWS Lambda, with another significant portion on hardware acceleration for cloud workloads (AI, databases, SDN).
What's exciting to me is that this is just the beginning. These students are representative of the next generation of developers. So much innovation left to go; I believe we'll eventually look upon the "cloud" as merely an extension of our own computers, auto-scaling and responding to any workload seamlessly much like how nowadays we rarely give second thought to memory management, garbage collection, etc. when writing in JavaScript or Python.
[1] https://stdlib.com/