Adding support and integration for SYCL compiler

[joshbainbridge]

We currently don't directly support the SYCL compiler, however it is a trivial one liner change to the CMake setup to get this to work:

diff --git a/CMakeLists.txt b/CMakeLists.txt
index ea337c0..855c411 100644
--- a/CMakeLists.txt
+++ b/CMakeLists.txt
@@ -55,7 +55,7 @@ endif()
if(${OPENQMC_ARCH_TYPE} STREQUAL GPU)
        set(CMAKE_EXPORT_COMPILE_COMMANDS OFF)
 
-       enable_language(CUDA)
+#      enable_language(CUDA)
 
        if(NOT DEFINED CMAKE_CUDA_ARCHITECTURES)
                set(CMAKE_CUDA_ARCHITECTURES OFF)

It would be great to provide support to as many platforms as possible. Adding SYCL appears to be fairly low hanging fruit we could action in the near term. It would be good to setup appropriate testing on the CI if this is added as a compiler option alongside the other target architectures.

I'd imagine this would look like separating the GPU option into two, CUDA and SYCL.