This event will be held remotely.
Registration deadline: January 30, 2021
SDSC supports the training of its user community, including students, in all aspects of high-performance computing (HPC). The goal of the training is to prepare new HPC users to run jobs on HPC systems. The sessions are designed to introduce participants to a variety of HPC topics and skill development activities. This years agenda includes:
Session #1: January 22, 2021 - March 12, 2021
- 1/22/2021: Program Orientation, history, plan ; Registration process & accounts; HPC overview & Expanse Architecture
- 1/29/2021: Parallel Queues & batch scripting; intro to applications (hello world, 2Djacobian)
- 2/5/2021:GPU Architectures, CUDA
- 2/12/2021:GPU Computing with CUDA Python
- 2/19/2021:CPU Computing: Introduction to OpenMP/Threads
- 2/26/2021:CPU Computing: Introduction to MPI
- 3/5/2021:CPU Profiling, monitoring: gprof, mpiP/comms, uProf
- 3/12/2021:GPU Profiling, monitoring: comms, NSight
-- Session Break (3/19/2021 & 3/26/2021)
Session #2: April 2, 2021 - May 7, 2021
- 4/2/2021: Containers for CPU and GPU computing
- 4/9/2021: Cloud Computing: AWS
- 4/16/2021: cloud computing: Azure
- 4/23/2021: Machine Learning; Tensor Flow
- 4/30/2021: SCC20 Applications: HPL, HPCG
- 5/7/2021: Accessing and Visualizing large data sets on HPC systems: using Jupyter Notebooks (pandas, to see Lustre data sets, Matplotlib, Boqeh, SeaBorn)
Topics are subject to change