take SYCL implementation from DPCT #4678

Job	Run time
clang-format	14s
linux_clang-14_cuda-11.2_debug_analysis	13m 36s
windows_cl-2022_debug_analysis	16m 11s
macos_xcode-14.2_debug_analysis	33m 18s
linux_gcc-12_debug_analysis	14m 22s
macos_xcode-13.2.1_debug	57s
macos_xcode-14.2_release	24m 59s
macos_xcode-14.3.1_debug	8m 16s
windows_cl-2022_release	20m 46s
windows_cl-2022_debug	15m 45s
linux_gcc-9_debug	15m 13s
linux_gcc-12_release_c++20	13m 30s
linux_gcc-13_debug	16m 20s
linux_clang-10_release	10m 3s
linux_clang-12_release	32m 20s
linux_clang-13_debug	7m 21s
linux_clang-16_debug_ubsan	10m 19s
linux_clang-16_debug_tsan	11m 33s
	4h 25m 3s

Provide feedback