lecture_slides [ECE1724S: Programming Massively Parallel Graphics Processors with CUDA]
Programming Massively Parallel Processors Using CUDA Site Home Page
Writing /fs1/eecg/moshovos/a/a3/moshovos/public_www/CUDA08/data/cache/8/8b871120725535ef08ad49f39bb58650.i failed
Unable to save cache file. Hint: disk full; file permissions; safe_mode setting.
Writing /fs1/eecg/moshovos/a/a3/moshovos/public_www/CUDA08/data/cache/8/8b871120725535ef08ad49f39bb58650.i failed
Unable to save cache file. Hint: disk full; file permissions; safe_mode setting.
Writing /fs1/eecg/moshovos/a/a3/moshovos/public_www/CUDA08/data/cache/8/8b871120725535ef08ad49f39bb58650.xhtml failed

Course Slides

Introduction to the Course

http://www.eecg.toronto.edu/~moshovos/CUDA08/slides/lecture-001.ppt (1.148 Mb) Overview of the CUDA execution model, and of a typical NVIDIA GPU containing system. Also, overview of the course.

Introduction to CUDA Programming

CUDA Hardware

Optimizing for CUDA

Reductions

Histograms and Sparse Matrix Multiplication

Scans and Streams

http://www.eecg.toronto.edu/~moshovos/CUDA08/slides/007 - Scans.ppt (1.708 Mb)

http://www.eecg.toronto.edu/~moshovos/CUDA08/slides/Scans Explained_1.ppt (211 Kb)

Added slides on explaining the balanced tree algorithm for calculating scans.

Textures

hiCUDA

Lecturer: Prof. Tarek Abdelrahman http://www.eecg.toronto.edu/~moshovos/CUDA08/slides/hicuda.pdf (1.273 Mb)

Profiler, Assembly, and Floating Point

 

lecture_slides.txt · Last modified: 2009/05/04 21:04 by yperxristis