In the second semester of my Master’s degree, I enrolled in the “Scalable High Performance Computing (SHPC)” course (along with GenAI).
The reasons I chose this course are…
- The lectures focus on GPU and Parallel Processing, which I find WakuWaku (exciting).
- I have enjoyed computing system courses since my undergraduate years.
- I believe learning to utilize GPU-based heterogeneous computing systems will be important for my future research papers.
Below is a lecture note I created based on the course content.
- SHPC 01 - Trends
- SHPC 02 - Binary Representations
- SHPC 03 - Floating Point Representations
- SHPC 04 - Processes and Threads
- SHPC 05 - Dependences and Pipelining
- SHPC 06 - Loop-Carried Dependences and Parallelism
- SHPC 07 - Synchronization
- SHPC 09 - GPU Architectures
- SHPC 10 - Caches
- SHPC 11 - Cache Coherence and Tiling
- SHPC 12 - Memory Consistency and Virtual Memory
- SHPC 13 - OpenMP
- SHPC 14 - CUDA
- SHPC 15 - MPI
- SHPC 16 - Register Allocation
- SHPC 17 - Multiple GPUs
- SHPC 17.5 - Multiple GPUs (CUDA)
- SHPC 18 - CUDA Streams
- SHPC 19 - Optimization for GPUs
- SHPC 20 - Shared memory
- SHPC 21 - AI parallelism