Skip to content

Updated README.md with single precision option for batched_linear_solver Sample #2142

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 2 commits into from Jan 22, 2024
Merged
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
11 changes: 10 additions & 1 deletion Libraries/oneMKL/batched_linear_solver/README.md
Original file line number Diff line number Diff line change
Expand Up @@ -42,7 +42,7 @@ Third party program Licenses can be found here: [third-party-programs.txt](https
>For more information on environment variables, see [Use the setvars Script for Linux or macOS](https://www.intel.com/content/www/us/en/develop/documentation/oneapi-programming-guide/top/oneapi-development-environment-setup/use-the-setvars-script-with-linux-or-macos.html).

### Running Samples on the DevCloud
When running a sample in the Intel DevCloud, remember that you must specify the compute node (CPU, GPU, FPGA) as well whether to run in batch or interactive mode. For more information see the Intel® oneAPI Base Toolkit Get Started Guide (https://devcloud.intel.com/oneapi/get-started/base-toolkit/).
When running a sample in the Intel DevCloud, remember that you must specify the compute node (CPU, GPU, FPGA) as well as whether to run in batch or interactive mode. For more information see the Intel® oneAPI Base Toolkit Get Started Guide (https://devcloud.intel.com/oneapi/get-started/base-toolkit/).

Run `make` to build and run the sample. Three programs are generated:

Expand All @@ -52,6 +52,15 @@ Run `make` to build and run the sample. Three programs are generated:

Note that the makefile only runs small tests to verify that the executables are working correctly. The problem sizes are too small to justify accelerator offload. Use the following command-line options to run the tests shown in [Solving Linear Systems Using oneMKL and OpenMP Target Offloading](https://www.intel.com/content/www/us/en/developer/articles/technical/solve-linear-systems-onemkl-openmp-target-offload.html): `-n 16000 -b 8 -r 1 -c 5`.

> **Note**: By default this makefile will be executable on devices with double precision support (fp64).
> To execute on devices with only single precision support(e.g., gen11, gen12) include `-DSP` option as follows in makefile
>
> ```
> IFX_OPTS_OFFLOAD = -DSP -DMKL_ILP64 -qopenmp -fopenmp-targets=spir64 -fsycl -L${MKLROOT}/lib/intel64 -lmkl_sycl -
>lmkl_intel_ilp64 -lmkl_intel_thread -lmkl_core -liomp5 -lpthread -ldl
> ```
>

You can remove all generated files with `make clean`.

### Example of Output
Expand Down