Really cool course. I'm enjoying scala as a language as it has many simple methods to achieve complex goals, including parallelization, however this course also presents parallelism in a rawer fashion explaining in depth some of the techniques to parallelize seemingly sequential-only tasks.

Last week's exercise is quite long, and albeit being a truly beautiful project, most of the implementation is not strictly related to parallel programming.

In any case a totally good course!

The course was well covered and very informative. As someone with zero experience in parallelism, I felt I could've absorbed a lot more if the topics were not completely new to me.

As mentioned in other reviews, the last assignment, Barnes Hut Simulation, was very time consuming and I found myself spending more time trying to understand the problem rather than implementing it. That was very challenging for someone with low academic background.

The course was ok, but at times too much theoretical when it comes to proofs of some methods are working as expected. I think its good to know the principles of parallelizing application and changing the potential algorithms, but this has applicability only in specific fields like comlex computations that need to be highly parallelized and less likely you would use these principles in standard backend development

There was some really nice material covered in this course. For me, this was the best one so far in the series. The instructors presented clear and understandable lectures. The week #4 programming task was a bit harder for me, but only because I kept forgetting to think simple and reflect on what had been covered in the course. Once I stepped away and came back everything fell into place.

Great explanations and insights around parallel programming. The instructors help you understand how parallel programming works and are doing great in making the complex topic understandable. To improve it further, there could be more explanations on performance differences between alternative solutions as well as some guidance on how to apply these ideas in practical Scala programming.

Awesome course to learn parallel programming and advantage of scala to achieve this.

Learned how to think algorithms in parallel being careful that it's not always better to do everything in parallel.

The only thing I didn't like that much is that the course has too many mathematical explanations and maybe that lectures should be optional for those that want to go further.

This course had a lot of depth to it, both in the lectures and the programming assignments. The last assignment seemed needlessly complex with not enough focus on the parallel programming aspect. I wish there had been more of a focus on how to actually write the parallel and task constructs which were instead given already implemented without much explanation.

Too difficult, especially the last week, especially for person without strong math background. I felt that material from the two last weeks is important and necessary to grasp PP, but speed should be slower.

I think the best way to improve this course would be splitting 3 and 4 weeks to obtain 6 weeks course like the first one in the specialization.

The stuff was interesting, but the lectures in the second week are monotonous. In the assignment in the fourth week one learns some interesting things about numerics for physical approximation only were little about parallel programming. Basically one just wants to make the assignment more challenging by letting you do some geometry.

On the plus side:

The material is presented in concise and efficient manner. It required little knowledge of Scala and JVM. It kept hidden the underlying implementation of parallelism and task utilities.

On the minus side:

The homeworks were easy. I could fill in the gaps almost automatically and I don't feel like I know the matter.

I have one issue about week 4 "Append" tree. I did not get how Append correlates with binary digits. I would like to have some step by step example how will we create a tree like s1 = Single(1) appendLeaf Single(2), s2 = s1 appendLeaf Single(3), s3 = s2 appendLeaf Single(4), s4 = s4 appendLeaf Single(5) using binary digits?

The course got me a bit lost on the most theoretical parts about folding, because the explanations and the whole concept take a while to arrive to a completion, being quite complicated. I had to re-run over those a few times. Nevertheless, all the comparison between aggregators and lego blocks help quite a lot

I find the contents (theory behind) more challenging to follow than previous courses on the specialization, the nice thing is that the practical exercises are built so that they are still manageable and a good help to understand the lessons. I am missing Scala's actor model in this course, why is not there?

The theory is superb and both instructors are experts in the field, you likely won't find many of the course's concepts online easily. The assignments could've focused a lot more on implementing parallelism itself, rather than implementing algorithms which take advantage of already implemented parallelism.

The course achieves its objectives, but the assignments require too much effort for solving general problems, and relatively little for parallelism-related tasks. In other words, understanding what the problem is takes far more time than solving it. The Lego-assisted teaching is a definite plus though!

Would be good to introduce more about Scala native solutions like Futures, parallel data structures and best ways of using them on practice, cases and best practices how to work with shared data, avoid deadlocks, how to work best when you have to process large data or when you have massive calculations.

It would be great to have some tests for understanding algorithms (like in algorithms course), not only practice programming assignments. Assignments are really interesting but can't verify your understand material or not. Also it is possible to pass assignment event if you have mistakes in the code.

Some missed explanations in the first week (mergedSort implementation)

Tha last two videos are quite difficult to understand. The ConcBuffer structure seems to be poweful, but the explanation and examples are tight.

In this course, the weekly tasks are explained very well. Thank you very much!

It requires more effort on undestanding the algorithms involved in the problems, rather than paralelism and its techniques. Instead of 4 weeks it would benefit of an additional week - for those who take the course after the regular work hours at their job. So 5 weeks would be more adequate.

I feel the concepts which are necessary to use in programming assignments are not discussed in the lectures enough. As result figuring out what the assignment steps are all about takes too much time. Some tasks are absolutely trivial, and others are simply too difficult.

There are some really boring parts, like 10 first minutes of many videolectures. It's kinda difficult to watch list.map explained in detail, since it's a course supposed to be taken after the one on basic Scala.

Overall - it's really good, I learned a lot.

Interesting course, very useful to get to know concepts of parallel programming. Assignments could be a little bit simpler but involving more programming from scratch, not just filling in methods bodies. This would help more to understand concepts.

Some problems with downloads and to view videos on the mobile app (Android), with some videos recorded with a low volume.

Despite that interesting and I think would need more time to get deeper in the concepts showed.

I learned very useful programming techniques, but I'm a bit lost as to whether these techniques are "good practices" or if they are "existing parallel programming features in Java, transposed as-is in Scala"

Appreciate the time spent by the authors to prepare all the course materials and the assignments. Everything clear and relatively self-contained. However, I still do not feel to be an expert in parallelism.