[SYCL] updates to oneapi compiler documentation

docs/polaris/compiling-and-linking/oneapi-compiler.md

@@ -1,36 +1,34 @@
-# OneAPI Compilers and Support
+# oneAPI Compilers and Support
 
-The Intel OneAPI compiler and Codeplay plugins for Nvidia GPUs are available on Polaris.
+The Intel oneAPI compiler and Codeplay plugins for Nvidia GPUs are available on Polaris.
 The oneAPI compilers are not enabled under the Cray Programming Environment system but can be used separately.
 Two oneAPI variants are provided, the first being a "release" version based on Intel's officially released oneAPI toolkit.
 [Intel Release Notes](https://www.intel.com/content/www/us/en/developer/articles/release-notes/intel-oneapi-toolkit-release-notes.html)
 
 !!! note
 
-    The 2023.1 release of oneAPI Toolkit does not support oneMKL or oneDPL on Nvidia.
+    The 2023.2.0 release of oneAPI Toolkit does not yet support oneMKL or oneDPL on Nvidia devices.
 
-The other variant being a build of main from the open source. This variant will be more up-to-date at the risk of bugs and breakages based on code that has not undergone a full release cycle.
-The documentation is located on the [SYCL](../programming-models/sycl-polaris.md) page.
+The other variant being a build from the open-source. This variant will be more up-to-date at the risk of bugs and breakages based on code that has not undergone a full release cycle.
+The documentation is located on the [SYCL](../programming-models/sycl-polaris.md) page. Most notable differences being, `icx/icpx` are the names of C/C++ compilers respectively when using the release version of the module where as `clang/clang++` are for open-source variant.
 
 
 ## Compile and Link
-OneAPI uses the clang (or icx/icpx wrapper) for compiling and linking for the Nvidia A100 SM80 architecture.
+oneAPI uses the clang (or icx/icpx wrapper) for compiling and linking for the Nvidia A100 SM80 architecture.
 
 
 ```
-module load PrgEnv-nvhpc
-module use /soft/compilers/oneapi/release/modulefiles
-module load compiler
-icpx -std=c++17 -fsycl -fsycl-targets=nvptx64-nvidia-cuda  -Xsycl-target-backend --cuda-gpu-arch=sm_80 test.cpp
+module load oneapi/release
+icpx -std=c++17 -fsycl -fsycl-targets=nvptx64-nvidia-cuda -Xsycl-target-backend --cuda-gpu-arch=sm_80 test.cpp
 ```
 
 ```
 harms@polaris-login-04:~/working/polaris/oneapi> icpx --version
-Intel(R) oneAPI DPC++/C++ Compiler 2023.1.0 (2023.1.0.20230320)
+Intel(R) oneAPI DPC++/C++ Compiler 2023.2.0 (2023.2.0.20230721)
 Target: x86_64-unknown-linux-gnu
 Thread model: posix
-InstalledDir: /soft/compilers/oneapi/release/2023.1/compiler/2023.1.0/linux/bin-llvm
-Configuration file: /soft/compilers/oneapi/release/2023.1/compiler/2023.1.0/linux/bin-llvm/../bin/icpx.cfg
+InstalledDir: /soft/compilers/oneapi/release/2023.2/compiler/2023.2.1/linux/bin-llvm
+Configuration file: /soft/compilers/oneapi/release/2023.2/compiler/2023.2.1/linux/bin-llvm/../bin/icpx.cfg
 ```
 
 ## Running
@@ -48,13 +46,13 @@ Expected output of sycl-ls and which platforms are available.
 
 ```
 harms@x3004c0s7b0n0:~> which sycl-ls
-/soft/compilers/oneapi/release/2023.1/compiler/2023.1.0/linux/bin/sycl-ls
+/soft/compilers/oneapi/release/2023.2/compiler/2023.2.1/linux/bin/sycl-ls
 
 harms@x3004c0s7b0n0:~> sycl-ls
-[opencl:acc:0] Intel(R) FPGA Emulation Platform for OpenCL(TM), Intel(R) FPGA Emulation Device 1.2 [2023.15.3.0.20_160000]
-[opencl:cpu:1] Intel(R) OpenCL, AMD EPYC 7543P 32-Core Processor                3.0 [2023.15.3.0.20_160000]
-[ext_oneapi_cuda:gpu:0] NVIDIA CUDA BACKEND, NVIDIA A100-SXM4-40GB 0.0 [CUDA 11.4]
-[ext_oneapi_cuda:gpu:1] NVIDIA CUDA BACKEND, NVIDIA A100-SXM4-40GB 0.0 [CUDA 11.4]
-[ext_oneapi_cuda:gpu:2] NVIDIA CUDA BACKEND, NVIDIA A100-SXM4-40GB 0.0 [CUDA 11.4]
-[ext_oneapi_cuda:gpu:3] NVIDIA CUDA BACKEND, NVIDIA A100-SXM4-40GB 0.0 [CUDA 11.4]
+[opencl:acc:0] Intel(R) FPGA Emulation Platform for OpenCL(TM), Intel(R) FPGA Emulation Device 1.2 [2023.16.7.0.21_160000]
+[opencl:cpu:1] Intel(R) OpenCL, AMD EPYC 7543P 32-Core Processor                3.0 [2023.16.7.0.21_160000]
+[ext_oneapi_cuda:gpu:0] NVIDIA CUDA BACKEND, NVIDIA A100-SXM4-40GB 8.8 [CUDA 11.4]
+[ext_oneapi_cuda:gpu:1] NVIDIA CUDA BACKEND, NVIDIA A100-SXM4-40GB 8.8 [CUDA 11.4]
+[ext_oneapi_cuda:gpu:2] NVIDIA CUDA BACKEND, NVIDIA A100-SXM4-40GB 8.8 [CUDA 11.4]
+[ext_oneapi_cuda:gpu:3] NVIDIA CUDA BACKEND, NVIDIA A100-SXM4-40GB 8.8 [CUDA 11.4]
 ```

docs/polaris/programming-models/sycl-polaris.md

@@ -10,14 +10,26 @@
 module load oneapi
 ```
 
-:warning: This module (compilers, libraries) gets built periodically from the latest open-source rather than releases. As such, these compilers will get new features and updates quickly that may break on occasion.
+!!! note
+
+This module (compilers, libraries) gets built periodically from the latest open-source rather than releases. For more details on the release version of compiler, please find the details [here](../compiling-and-linking/oneapi-compiler.md). As such, these compilers will get new features and updates quickly that may break on occasion. Please submit any issues at the respective github repositories for the compilers and libraries.
+
+## Components
+- These are the list of components associated with this module
+
+| User Application  | Component                                               |
+|-------------------|---------------------------------------------------------|
+| Compilers         | [DPC++](https://docs.nvidia.com/cuda/cublas/index.html) |
+| oneMKL Interfaces | [oneMKL](https://github.com/oneapi-src/oneMKL)          |
+| oneDPL            | [oneDPL](https://github.com/oneapi-src/onedpl)          |
+| SYCLomatic/DPCT   | [dpct](https://github.com/oneapi-src/syclomatic)        |
 
 ## Dependencies
 - SYCL programming model is supported through `oneapi` compilers that were built from source-code
 - Loading this module switches the default programming environment to GNU and with the following dependencies
   - PrgEnv-gnu
   - cudatoolkit-standalone
-- Environment Variable set: `SYCL_DEVICE_SELECTOR=ext_oneapi_cuda:gpu`
+- Environment variable is set when loading the module: `ONEAPI_DEVICE_SELECTOR=ext_oneapi_cuda:gpu`
 
 ## Example (memory intilization)
 
@@ -139,13 +151,13 @@ For further details regarding the arguments passed to `mpiexec` command shown ab
 [oneMKL Interfaces](https://github.com/oneapi-src/oneMKL) is an open-source implementation of the oneMKL Data Parallel C++ (DPC++) interface according to the [oneMKL specification](https://spec.oneapi.io/versions/latest/elements/oneMKL/source/index.html). It works with multiple devices (backends) using device-specific libraries underneath.
 
 oneMKL is part of oneAPI. Various backend supported are shown below. More Information [here](https://github.com/oneapi-src/oneMKL#supported-configurations).
+
 | User Application | Third-Party Library                                          |
 |------------------|--------------------------------------------------------------|
 |                  | [cuBLAS](https://docs.nvidia.com/cuda/cublas/index.html)     |
 | oneMKL interface | [cuSOLVER](https://docs.nvidia.com/cuda/cusolver/index.html) |
 |                  | [cuRAND](https://docs.nvidia.com/cuda/curand/index.html)     |
 
-
 ## Example (using onemkl::gemm)
 The following snippet shows how to compile and run a SYCL code with oneMKL library. For instance, a GPU-based GEMM is performed using `mkl::gemm` API and the results are compared to a CPU-based GEMM performed using the traditional blas (e.g., AOCL-BLIS) library.
 ```c++
@@ -272,6 +284,7 @@ int main() {
 ```
 
 Compile and Run
+
 The user would need to provide paths the math-libraris as shown below. Also please provide AOCL library for CPU GEMM by `module load aocl`.
 Environment variables `MKLROOT` is defined with `oneapi` module & `AOCL_ROOT` is defined with `aocl` module.
 Note: Please pay attention to the linker options for AOCL & oneMKL libraries.