Ginkgo’s SYCL backend and its integration into nekRS

The increasing diversity of hardware architectures makes it harder for simulation software to get everything working on different hardware. Simulation software stacks need to rely on portability layers or software stacks that support different hardware architectures. We present Ginkgo, a math library that considers platform portability seriously and can be used for linear solvers in nekRS CFD code. We have presented nekRS with Ginkgo on AMD and NVIDIA GPUs. With the next release of nekRS supporting Intel GPUs, we also extend the Ginkgo integration to keep the portability on Intel GPUs. In this talk, we will show performance results from nekRS simulation with Ginkgo solver on Intel GPUs.