A brief write-up on how to get llama.cpp working with SYCL via Intel's oneAPI for Intel Panther Lake processors with Arc iGPUs on Gentoo Linux.