Aurora Learning Paths: Migrating from CUDA to SYCL

CUDA to SYCL migration

SYCL is becoming a de facto standard for vendor-agnostic heterogeneous computing. Upgrading CUDA code to standard C++ with SYCL makes the applications portable across various existing and evolving accelerators including Intel and Nvidia GPUs.

The "Migrating to SYCL" portal on Intel Developer Zone is designed to help the worldwide developer community successfully port their applications to oneAPI, helping developers break free from propriety APIs and provide greater customer choice in hardware for performance, power usage, and cost. This portal currently features 1) A step-by-step approach based on the tools provided by Intel to achieve the portability and performance goals. 2) End-to-end migrated solutions for three real-world examples. 3) Migration examples from researchers and engineers across the world. Developers are encouraged to take advantage of this portal, migrate their CUDA code to SYCL and publish on DevMesh, Intel’s community for developers and creators.

Click here to register

Chekuri S. Choudary, Intel Program Manager

Chekuri S. Choudary joined Intel as the technical lead and program manager for the Migrating to SYCL program. Prior to joining Intel, he worked with several clients and universities across the globe evangelizing AI/HPC and designing solutions by integrating products with open-source software. He was an applied researcher in AI and high-performance computing collaborating with U.S. DOE labs, NASA, and universities on SBIR projects. 

Rakshith Krishnappa, Intel Developer Evangelist
Rakshith is a Developer Evangelist at Intel with over 17 years of experience in software development and Intel products. He graduated from the Illinois Institute of Technology with a master’s in Electrical and Computer Engineering.  His focus at Intel is in high-performance computing and oneAPI Products and Solutions.