Introduction to Parallel Programming

Sign up for the class here:

Learn the fundamentals of parallel computing with the GPU and the CUDA programming environment! In this class, you'll learn about parallel programming by coding a series of image processing algorithms, such as you might find in Photoshop or Instagram. You'll be able to program and run your assignments on high-end GPUs, even if you don't own one yourself.

[beenviolent]

"How long will the course be available?
This class will always be available"
Oof.

[daviangel]

Yes this makes a lot more sense because I've programmed using both CUDA and OpenCL and CUDA isn't much easier to program! I guess if you take into consideration better tools integration with Visual Studio, debugging support, etc it might be a bit easier but not by much. They are both still pretty low level and require a lot of boilerplate code to get up and running. Coursera course is also using Cluster GPU Instances for Amazon EC2 for student programming although OpenCL is also supported.

[johnowens]

This is my first Udacity course too, so you might just browse some other courses to see generally how they work. I believe generally they have lectures with interspersed quizzes and projects or hw assignments (we have projects). As for background, "solid experience with the C programming language and basic knowledge of data structures and algorithms" is a good summary, I think.

[johnowens]

One of my students noted today that one of the benefits of OpenCL is particularly excellent integration with Visual Studio. Gotta say, though, that when I look at an OpenCL "hello world" (search on, say, OpenCL Hello World Example and look at the Apple example), it seems considerably more complicated than a CUDA hello world.

[johnowens]

We've chosen to put our energy/class time into CUDA. I agree there are a significant number of interesting other languages/environments. If I picked another one, it'd probably be OpenACC. I am honestly a little wary of teaching something that would be both single-vendor AND single-platform, although C++ AMP might grow away from either of those in the future, which would make it more interesting for teaching purposes. (It'd be near-impossible to teach C++ AMP in the cloud, for instance.)

[MrShebaw]

Thank you for taking time to reply. We can't wait for this class. It's going to be awesome.

[Tannzrz]

Suggestion: briefly go over C++ AMP as well. It's beyond simple, especially if your students catch onto CUDA.

[daviangel]

Cool Seymour Cray quote! never heard that one before but great analogy

[johnowens]

So I discussed this a bit on the reddit page for this course (go search for "cs344" on reddit)—I'll be happy to discuss this in more depth there. Briefly, though: CUDA is simpler and has a more developed ecosystem. Good news, though, is that if you know one of {CUDA, OpenCL}, it's quite easy to pick up the other, because the programming models are so similar.

[daviangel]

You definitely need to have a good grasp of C and math to include Linear Algebra. I"m taking the Coursera parallel class and there is a lot of both.

[johnowens]

Background: "solid experience with the C programming language and basic knowledge of data structures and algorithms". Like other Udacity classes, we have microquizzes during the lectures and a significantly-sized programming assignment (that you'll do in an online IDE) per lecture.

[johnowens]

Yes, programming assignments are in C (with CUDA).

[MalamIbnMalam]

Why did they remove this course on Udacity?

[johnowens]

Ooh! Do tell, what's the annoying orange? My students will be quite interested.

[johnowens]

Good news! All the programming assignments in the class will be done on GPUs in the cloud (through a web-based IDE). We want the class to be accessible to as many people as possible, even those with no GPUs at all.

[zohaibisrunning]

Why not OpenCL ? Is not it better to switch to a vendor independent technology than staying specific to CUDA?

[czmelak]

on the contrary, that's a very good analogy.

[SaiHanumanthaRao]

The course is not available, is it moved some where?

[Tannzrz]

It appears that making it multi-platform is their intention, see:

But I understand, since it's still a work-in-progress. Thanks for the heads-up!

[deimos]

So, are we going to use C in our coding assignments?