oneAPI Workshop 051523
Published 2023-05-26
C++ SYCL provides a consistent programming language across CPU, GPU, FPGA, and AI accelerators in a heterogeneous framework where each architecture can be programmed and used either in isolation or together.
The language and API extensions in SYCL enable different development use cases, including development of new offload acceleration or heterogeneous compute applications, conversion of existing C or C++ code to code that's compatible with SYCL, and migrating from other accelerator languages like CUDA to SYCL.
Learn basics of C++ SYCL programming to offload computations to accelerator, Intel Offload Advisor to assist in offloading computations to accelerators, SYCLomatic Tool to migrate CUDA source to SYCL source and Intel VTune Profiler to help in performance optimization.
Agenda
10:00AM – 10:30AM
Brief Overview of Portability of Amber to SYCL project
Andreas Goetz, PhD, Research Scientist and oneAPI Center-of-Excellence PI, SDSC/UCSD
10:30AM – 11:00AM
C++ to SYCL - Assuming knowledge of basic C++, we will introduce SYCL basics for offloading computation to GPU
Rakshith, Intel
11:00AM – 12:30PM
SYCL Essentials - SYCL Program Structure, SYCL Memory Models and Overview of Kernel Programming
Rakshith, Intel
12:30PM – 1:00PM - Break
1:00PM – 1:30PM
CUDA to SYCL Migration using SYCLomatic Tool
Rakshith, Intel
1:30PM – 2:00PM
Intel Advisor and Intel Vtune Profiler tool overview
Rakshith, Intel