| Time & Place: | TR 9:30-10:45 PM; meeting in FPAT 265 |
|---|---|
| Instructor: | Professor Hank Dietz |
| Instructor URL: | http://aggregate.org/hankd/ |
| Course URL: | http://aggregate.org/GPUCOURSE/ |
Graphic Processor Computing is about the emerging many-core parallel computing architecture that has grown out of hardware originally developed as Graphics Processing Units (GPUs) for high-end video cards.
The first exam was in class on Thursday, October 11, 2018. Review for the exam was in class on Tuesday, October 9, 2018.
We will not have class on Tuesday, October 23, 2018 -- Professor Dietz will be presenting a paper at the 9th International Green And Sustainable Computing Conference.
All course materials will be linked here:
Aside from working on your own systems, there are two alternatives. There will soon be two NVIDIA GPU systems in Marksbury for this course, but we also might be using the large GPU system managed by CCS.
Assignment 0: How Fast Is This? is now posted. This assignment must be done using one of the two systems we are making available. It is due before class, Thursday, October 4, 2018.
Assignment 1: Traveling Salesperson Genetic Algorithm is now posted.
This assignment must be done using one of the two systems
we are making available.
It was origininally to be due before class, Thursday, October 18, 2018,
but the deadline has been extended by one week to October 25...
and now it has been extended one more week, to November 1.
It will not be extended again.
Here's a little example of doing island
model communication in the host.
Assignment 2: Nearly Perfect Hash Functions is now posted (and the forms file format is now much simpler than we originally discussed). This OpenACC project is due before the final exam timeslot. That means before 10:30AM on Tuesday, December 11, 2018.
Professor Hank Dietz is usually in the Davis Marksbury Building; see his home page for
complete contact info. He has an "open-door" policy that
whenever his door is open and he's not busy with someone else,
he's available -- and yup, there really is a slow-update live
camera in his office so you can check.
Alternatively, you also can email hankd@engr.uky.edu to make an appointment; please
use "GPUCOURSE" in the email subject line for anything
related to this course.
About the graphic: The first GPU that was truly intended as a computing device, rather than a graphics card, was the ATI FireStream, shown in the fisheye image. Since then, ATI (now part of AMD) and NVIDIA have been the major players, with NVIDIA's CUDA software environment dominating the field. OpenCL is the newer "generic" software environment being pushed as a portable standard. There's even some significance to the color and placement of the text... but I've said enough for now. ;-)
GPU Computing
