Hide menu

TDDD56 Multicore and GPU Programming (6 ECTS)

Autumn2 2018

Mid-Term Evaluation with Muddy Cards

The course TDDD56 has been mid-term evaluated with the muddy card method in the third week, Lecture 7, on monday 13 november 2017 08:00-10:00. 17 students attended this lecture, and I received 17 cards.


By and large, the result is clearly positive. 7 cards were exclusively positive, no card was exclusively negative.

The course contents is perceived as interesting and challenging.
One card appreciates that the overlap of several (CPU) lectures with TDDC78 is clearly marked. (This overlap in some core topics is unavoidable in order to make both courses self-contained.)
One card mentions that not all students have a core computer science background and that some basic notions are not known ...
(Comment: based on that day's lecture topic, I can only guess that this remark refers to analysis of algorithms and to the big-O notation for asymptotic behavior of (time complexity) functions),
... and suggests to provide a list of key words to get familiar with before the start of the course.
(Comment: We will consider this for next year's TDDD56. For now, you can find concepts in process and thread programming in the slide set of Lecture 3. Basic algorithm analysis and Big-O notation should actually be known from any reasonable course in data structures and algorithms, which was also listed in the prerequisites. However, feel free to have a look at this slide set on the analysis of sequential algorithms and Big-O notation from a previous course on data structures and algorithms, hopefully it provides the desired background information.
If there are any further concepts that are unclear, please let me know.

At the time of the evaluation, most students have finished CPU Lab 1, some had already started with CPU lab 2.
Many cards praise the labs so far (challenging, interesting, right level of difficulty, well documented, helpful for understanding the theory, fun) and the lab assistants.
The smaller sizes of groups B and C are appreciated, also that Norrköping based students have the possibility to avoid late evening labs. One card says that it can be hard to get hold of the lab assistant when many people need help, and suggests a queue on the whiteboard.
A few cards find the labs tough or desire clearer lab instructions. 2 cards remark on typos and English grammar issues in the lab scripts resp. the lab instructions (esp. Lab 2).
One card complains on lots of confusing preprocessor usage in the lab skeletons.
One card appreciated that Lab 1 was not so difficult and a good starting point. One card finds that the final part of Lab 2 (how to force the ABA problem) felt pointless and could be replaced by questions when demoing to make sure that the concept is fully understood.
Several cards dislike the mandatory presence at the labs.
(Comment: Nobody needs to stay throughout the entire lab session once the solution has been successfully demonstrated. Even better, use the time and start already with the next lab if possible.)
One card appreciates that the CPU labs can be prepared on any POSIX computer.

Also the lectures and lessons get mostly positive comments.
Lectures are perceived as clear, interesting, well structured, well presented and coherent with the labs.
The slide material is, in general, appreciated. One card finds that there is much text but values the extensive slide material for exam preparation, while another card actually desires more text and self-explanation to help students who missed the lecture.
2 cards suggest more examples on the whiteboard.
The one lesson we had so far was appreciated as a good preparation for the labs.

Other aspects:
The course is perceived as well-structured and organized.
The course web page functionality is well appreciated.
One card suggests using LiSaM e.g. for restricted access to the lecture notes.
(Comment: LiSaM contains much functionality that we do not need for our courses because we already have (and always had) well functioning course web pages. The current .htaccess based solution for restricted access is quickly set up and is more flexible than LiSaM because it also works for non-registered students, e.g. remote and PhD students following the course and former students taking re-exams. We are however open to using selected LiSaM functionality in the future.)

Thanks to everybody for the overall very constructive and encouraging feedback!

Christoph Kessler, examinator TDDD56

Page responsible: Christoph W Kessler
Last updated: 2017-11-17