Short Courses of the 23rd Symposium on High Performance Computing Systems
Keywords:
High Performance Computing, WSCAD 2022 Short Courses, WSCAD 2022Synopsis
In this book are the six mini-courses presented during the 23rd Symposium on High Performance Computing Systems (WSCAD 2022), held between October 19 and 21, 2022 in Florianópolis, SC. In all mini-courses, the practical bias that encourages participants and readers to use high performance effectively stands out. Chapter 1 "OneAPI: A Developer-Centered Approach to Heterogeneous Computing" introduces the OneAPI standard, which is an open standard for computing acceleration that takes a single-source approach, in which all application code can be specified. Uniformly using the C++ language, regardless of running on host or accelerators. Chapter 2 "Artificial Intelligence and Function as a Service: Provisioning Applications with AWS Lambda" introduces AWS Lambda, which is one of the main tools of the so-called serverless computing service. Within this context, the concepts of Function as a Service (FaaS) are introduced, encompassing the use of AWS Lambda to provision AI applications. Chapter 3, "Things to Do Before Parallelizing", presents practices that should be done to improve code performance before considering parallelization, and reviews essential concepts for evaluating and comparing programs, including examples and live demonstrations of how to apply them. Chapter 4 "Fundamentals of Accelerated Computing with CUDA C/C++" teaches the fundamental tools and techniques to accelerate applications written in C/C++ languages to run on massively parallel architectures with CUDA. Chapter 5 "From sequence assembly to ancestry testing: HPC challenges for bioinformatics" provides an introductory overview of the process from obtaining to analyzing biological data, showing the main algorithms and database at each step and a discussion of ways to parallelize of these algorithms and the main challenges that still lack a solution. Chapter 6, "Tools for Configuring and Managing a High Performance Cluster in a Public Cloud", presents the best practices adopted by NACAD-COPPE/UFRJ for the implementation of a High Performance cluster using a Public Cloud, focusing on the use of tools for deploy and manage High Performance Computing (HPC) clusters on the AWS Cloud.
Chapters
-
1. OneAPI: A Developer-Centered Approach to Heterogeneous Computing
-
2. Artificial Intelligence and Function as a Service: Provisioning Applications with AWS Lambda
-
3. Things to Do Before Parallelizing
-
4. Fundamentals of Accelerated Computing with CUDA C/C++
-
5. From Sequence Assembly to Ancestry Testing: HPC Challenges for Bioinformatics
-
6. Tools for Configuring and Managing a High Performance Cluster in a Public Cloud
Downloads
