EE548 High Performance Computing

Lecture

Wednesday 13h00-15h00 (s5109)

Resources

C tutorial (https://www.learn-c.org)
CUDA toolkit documentation (https://docs.nvidia.com/cuda/index.html)
Introduction to CUDA (https://devblogs.nvidia.com/even-easier-introduction-cuda/)
CUDA unified memory (An Even Easier Introduction to CUDA)
MPI-2 Documentation (DeinoMPI)
MATLAB Parallel Computing Toolbox (Documentation API)

HPC server

The HPC server is accessible through SSH. Use the following instructions to connect using Visual Studio Code.

Schedule

Week	Date	Day	Lectures	Readings	Laboratory
1	11-15 Sep	Tue	Introduction to HPC Heterogeneous data parallel computing	Chap 1 & 2	https://www.learn-c.org Lab 0 (start code) (due 19 Sep)
2	18-22 Sep	Tue	Multidimensional grids and data Compute architecture and scheduling	Chap 3 & 4 (2D arrays in C)	Lab 1 (start code) (due 3 Oct)
3	24-29 Sep	Tue	Memory architecture and data locality Performance considerations	Chap 5 & 6
4	2-6 Oct	Tue	Convolution Stencil	Chap 7 & 8	Lab 2 (start code) (due 17 Oct)
5	9-13 Oct	Tue	Parallel histogram Reduction	Chap 9 & 10
6	16-20 Oct	Tue	Prefix sum (scan) Merge	Chap 11 & 12	Lab 3 (start code) (due 31 Oct)
7	23-27 Oct	Tue	Sorting Sparse matrix computation
8	30 Oct – 3 Nov	Tue	Graph traversal	Chap 13 & 14	Lab 4 (start code) (due 14 Nov)
9	6-10 Nov	Tue	OpenMP Intro to Thrust library	Barlas Read 8.1 to 8.3
10	13-17 Nov	Tue	Programming a heterogeneous computing cluster	Chap 20	Lab 5 (start code) (due 28 Nov)
11	20-24 Nov	Tue	Project proposal
12	27 Nov – 1 Dec	Tue	Project progress review
13	4-8 Dec	Tue	Project progress review
14	11-15 Dec	Tue	Project presentation		Project presentation Paper critique (both due 12 Dec)
15	18-22 Dec		Term is over