PAR Class 16, Mon 2020-03-23
Table of contents
1 About this unprecedented situation
RPI wants to continue to give you a solid education.
We recognize that new problems may arise and will try to be humane.
You will still have to study.
Anything listed here is only a best attempt and might be changed if there's a good reason.
Your feedback is always welcome.
2 Course computer tools
2.1 Webex
For the rest of the semester, the idea is that I'll still lecture at the appointed times, Mon and Thurs at noon eastern time.
However, now I'll try to lecture with Webex. You can access it via your favorite browser. Or, you can use an app on some platforms like the Ipad.
The lectures will be recorded so that you can rewatch them later.
I'll post the link on piazza with email to you before the lecture.
2.2 Piazza
I'll use piazza---
- to push announcements to the class,
- to post material that might not be completely public,
- for questions and answers.
2.3 This blog
This blog will continue to be sort of a permanent record of the class.
2.4 Gradescope
Gradescope will continue to be used for grading.
2.5 parallel.ecse
parallel.ecse will continue to be available remotely.
3 Today
3.1 Final round 2 student talks
Two remaining students will present your talks.
If possible, post your slides in advance or email them to me for posting.
3.2 OpenACC
Quickly finish the 3 slide sets.
From https://www.openacc.org/events/openacc-online-course-2018.
My local copy on parallel: /parallel-class/openacc/online-course
3.3 CUDA
I've mentioned this many times. Now we'll get into details.
- Lecture 5: Finish CUDA,
- Lectures 6, 7: Some parallel patterns.
- Sample programs.
- /local/cuda/samples has many programs by Nvidia. Some interesting ones:
- /local/cuda/samples/1_Utilities/deviceQuery/deviceQuery describes the host's GPUs.
- /local/cuda/samples/0_Simple/ does some benchmarks.
- Stanfords examples are in parallel-class/stanford/tutorials .
- /local/cuda/samples has many programs by Nvidia. Some interesting ones: