About the Course/Instructor

CSC 213 - Operating Systems and Parallel Algorthms - Professor Jerod Weinman

1  About the Course

Q:
When will we be using the NVIDIA GPUs? And how much parallel programming will we be doing on both the GPUs, and standard C on the old fashioned CPU?
A:
We'll do one lab in weeks 10-12 on the GPU. Many other labs will be parallelized on the multicore CPUs.
Q:
Also, I was wondering if you would address the average or expected length (wrt to time spent) for laboratory assignments in the class. I'm sure it is variable, as there are labs with multiple parts, but I was just curious (if there's even a meaningful answer).
A:
This is a great question. While I hope the coding of the assignments are straightforward, debugging race conditions can be tricky and will require careful thought, discussion with your partner, and perhaps prompt interaction with the mentor or the instructor. Write-ups are not designed to be big ordeals, but practicing the discipline by explaining empirical results is just as important as writing code. (That didn't answer your question directlly-I welcome statements of the form "We spent X amount of time on this lab" in your submissions.)
Q:
Will they be holding review sessions prior to the exams and the final?
A:
Good question! I bet that can be arranged.
Q:
Do we need to write a very small and basic OS for this class?
A:
We won't be writing an entire OS. You'll write a shell, and otherwise deploy your understanding of OS concepts.
Q:
What will the non lab homework assignments look like?
A:
Reading journals.
Q:
Can you give a couple example topics for Significant Bits talks?
A:
Follow the links given on the SigBits page - there are systems developments happening every day! Some of the talks from the last time I taught the course (four years ago), were the debut of Google's chrome browser (and how it handled threading versus processes), GPGPU computing (one of my favorites from that semester due to its cleverness), ZFS, chaotic logic gates, a new branch prediction technique, etc..

2  About Me

Q:
What are some hobbies you have? And what do you like to do in your free time(if it's not too personal)?
A:
I barbecue whenever I can (including Turkish!). I've started reading a lot of fiction as of February before last. I usually read a novel every week or two. (This week I finished Robert Harris's The Fear Index, which was not nearly as good as Neal Stephenson's REAMDE.) I don't usually read Sci-Fi, but I like a good thriller and other novels that take me to new places. I also like to fish from my kayak or small motorboat (both are 14').
Q:
As a computer science student, should I focus more on academic stuff like algorithms, or more applied things like application development? (Its more like a life question.)
A:
One thing Grinnell does well is teach you the fundamentals of computing very well (i.e., algorithms and theory), while also giving you some practical design experience (i.e., in 323 and 325). Your critical thinking and writing skills will enable you to learn specifics of application development, which are changing constantly.
Q:
You were a visiting professor at another college this past year, right? Where was that and what cool things were you doing?
A:
Yup! I was at UMass Amherst doing vision research, submitted a journal article on scene text recognition and continued to develop my project on making historical map images searchable for scholars (or the curious).
Q:
I dont have any questions pertaining directly to the course, but I am interested in knowing what got you interested in computer science.
A:
Mostly I'm a guy who likes constructive, logical thinking, solving problems, and building things. There's a certain amount of creativity to the whole thing to, that helps keep me balanced.
Q:
And do you know the gender of your baby yet?
A:
Nope.
Q:
What is one of the more important ethical questions to be asked of/about the field of artificial intelligence?
A:
Oooh! I love that question. I teach a whole tutorial on it. I think the question of moral responsibility in agency is fundamental. This summer Sebastian Thrun was asked (at CVPR 2012) "Who's responsible when a Google self-driving car crashes?" I was horribly disappointed by his (non)answer. "We just have to engineer them well enough so they don't." (A paraphrase, but not far from his original remarks.)
Q:
Why do the colored girls go doo do doo do doo do do doo?
A:
Probably ought to ask Mr. Reed.