[SlySherZ]

For everyone interested in this, I would like to recommend the MIT Advanced Data Structures course, by prof. Erik Demaine (mentioned in the article). The part of the course about cache oblivious data structures starts on this lecture:

https://www.youtube.com/watch?v=V3omVLzI0WE

[mbil]

Great suggestion -- Demaine's lecture videos were so illuminating for me when I studied cache-oblivious algos. Here's a corresponding paper describing the cache-oblivious Funnelsort algorithm: http://erikdemaine.org/papers/BRICS2002/paper.pdf

[utopcell]

I'd like to second that, Demaine is an awesome teacher.

[quantum-qaem]

I watched the lectures last week but it was difficult to understand. Can you recommend any beginner data structures course?

[SlySherZ]

As @Cursuviam said, Introduction to Algorithms is a pretty good introduction:

https://www.youtube.com/playlist?list=PLUl4u3cNGP61Oq3tWYp6V...

There are two things that make the lectures difficult to follow, IMO:

The first thing is that you need to be comfortable with the names that are given to things, understand well the simpler data structures like different kinds of trees, linked lists, graphs and so on, with their corresponding operation running times. Go watch Algorithms and you'll be fine :)

The second is that the subject is really hard. During the course, most of the time we're learning about (close to) state of the art data structures, so don't expect to understand everything in one go.

If you want to understand it in depth, you'll probably have to more time studying each lecture. Try thinking about how you'd solve the problem, try to understand the proposed solution and go through the video very slowly making sure you understand the details, try to implement at least some of them.

I used these lectures as a way to get comfortable with the field and get a broad overview of the kind of solutions that usually work, so I still have to go through most of Step 2.

[Cursuviam]

Why not Introduction to Algorithms? (also taught by demaine)