Instructor: Maria
Pantoja
California Polytechnic State University, CA, USA
Instructor: Erik
Pautsch
Loyola University Chicago, IL, USA
Instructor: Silvio
Rizzi
Argonne National Laboratory, Lemont, IL USA
Instructor: George K.
Thiruvathukal
Loyola University Chicago & Argonne National
Laboratory, IL, USA
Instructor: Alvaro
Vazquez-Mayagoitia
Argonne National Laboratory, IL, USA
Instructor: Raymundo
Hernandez-Esparza
Argonne National Laboratory, IL, USA
Brief Program:
- Presentation
- Introduction
- Heterogeneous computing and SYCL
- Hands-on Tutorial (1)
- Evaluation of Electron Density in Real Space Grid with SYCL
- Hands-on Tutorial (2)
- Porting the Marching Cubes visualization algorithm from CUDA to SYCL
- Hands-on Tutorial (3)
- Implementing an N-body simulation in SYCL
- Wrap up and closing remarks
Accelerating Computing Using SYCL Programming for GPUs
Information
In this tutorial, we will provide a concise introduction to SYCL, a parallel programming model designed to develop portable C++ code. With SYCL, developers can efficiently offload computations to accelerators, particularly GPUs. We will cover the fundamental concepts of GPU development, including a comprehensive walkthrough of exemplary cases. Additionally, we will discuss the limitations and strategies for migrating existing GPU codes to SYCL.
Student´s prerequisites
The intended audience is intermediate students in computer science with
knowledge of C/C++ programming. We can do it in English or Spanish,
depending on the preference of the
audience.
Materials
Students should create an account here: https://console.cloud.intel.com/
Attendees
will require a laptop computer that can view and run jupyter notebook
specifically
https://console.cloud.intel.com/
Previous
editions
Carla 2023 edition
References
https://unoapi.org/index.html
https://www.khronos.org/sycl/
https://github.com/codeplaysoftware/syclacademy
More information: T05-Maria-Pantoka-SYCL-v2