Events

About the workshop and live demo:

In this session, participants will explore how to program High-Performance FPGA cards using the Intel DPC++ compiler, a SYCL implementation, to facilitate heterogeneous device programming.

Field-Programmable Gate Arrays (FPGAs) are integrated circuits that can be configured by users post-manufacturing. They consist of a network of programmable logic blocks and reconfigurable interconnects, enabling the creation of custom digital circuits. Traditionally, FPGAs have been programmed using Hardware Description Languages (HDLs) such as Verilog and VHDL. However, designing high-performance accelerators with these languages demands extensive knowledge and experience in hardware design.

By leveraging higher-level abstractions like SYCL and OpenCL—C/C++-based programming models familiar to software developers—developers can generate hardware kernels through an offline compiler. This significantly simplifies FPGA programming, reducing development time compared to HDL, which typically involves more complex coding, simulation, and debugging processes.

Following an introduction to Meluxina’s FPGA cards and an overview of SYCL programming using the OneAPI software development toolkit, participants will engage in a live demonstration led by the presenter. This demonstration will feature the coding of a brief example: “How to code quantum circuits on FPGA”.

This session is open to every FPGA enthusiast and person interested in learning from SYCL programming.

About the speaker:

Emmanuel Kieffer:

Emmanuel Kieffer is an HPC (High-Performance Computing) Software Engineer in the Supercomputing Applications Team at LuxProvide, where he contributes to the development and optimization of software for MeluXina, the EuroHPC supercomputer based in Bissen, Luxembourg.

Before joining LuxProvide, Emmanuel was an HPC Research Scientist at the University of Luxembourg, focusing on large-scale optimization strategies leveraging high-performance computing.

His expertise spans heterogeneous programming, code optimization, and software acceleration, with a strong emphasis on improving computational performance and efficiency in HPC environments.