Schedule
Each talk is of 90 minutes duration.
Lecture | Date | Time | Topic | Remarks |
1 | May 2 | 17:00 | Introduction to GPU | |
2 | May 4 | 17:00 | Hello World in CUDA | |
3 | May 6 | 17:00 | CUDA program flow and cudamemcopies | |
4 | May 9 | 17:00 | Thread organization and Thread block | |
5 | May 11 | 17:00 | GPU computation Hierarchy and Matrix Squaring | |
6 | May 13 | 17:00 | Thread Divergence | |
7 | May 16 | 17:00 | Types of memories, latency, and locality | |
8 | May 18 | 17:00 | Memory coalescing and AoS vs. SoA | |
9 | May 20 | 17:00 | Shared Memory | |
10 | May 23 | 17:00 | Bank Conflicts | |
11 | May 25 | 17:00 | Compute Capabilities | |
12 | May 27 | 17:00 | Data Race and Synchronization | |
13 | May 30 | 17:00 | Mutual Exclusion | |
14 | June 1 | 17:00 | Atomics | |
15 | June 3 | 17:00 | Barriers | |
16 | June 6 | 17:00 | Reductions and Prefix sum | |
17 | June 8 | 17:00 | Functions | |
18 | June 10 | 17:00 | Support | |
19 | June 13 | 17:00 | CUDA Topics | |
20 | June 15 | 17:00 | Multi-dimensional arrays in C/CUDA | |
21 | June 17 | 17:00 | Exam (online, one hour, MCQ) | |
Certificate
A course completion e-certificate will be issued to all the registered participants who attend at least 15 of the classes live and write the exam. The exam score will
not be considered for certification.
Instructor
Jash Khatri
MS Scholar, CSE, IIT Madras
Researcher on Graph Algorithms on GPUs
(to be) Software Engineer at Microsoft
|
|