Skip to content
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

Docs: mention FPGA backend tested with Intel Quartus PRO #1335

Merged
merged 4 commits into from
Aug 2, 2023
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
10 changes: 9 additions & 1 deletion doc/setup/installation.rst
Original file line number Diff line number Diff line change
Expand Up @@ -21,7 +21,9 @@ however, it requires two more runtime dependencies to be installed and available
**GPU**: For NVIDIA GPUs, the CUDA toolkit is also required, and AMD GPUs require HIP. :ref:`See more information on how to configure DaCe to use AMD GPUs <amd>`.
You may (optionally) want to install `CuPy <https://cupy.dev/>`_ for easy integration of GPU arrays in Python.

**FPGA**: Xilinx FPGAs require the Vitis suite and Intel FPGAs require the Intel FPGA SDK to be installed.
**FPGA**: Xilinx FPGAs require the Vitis suite and Intel FPGAs require the Intel FPGA SDK to be installed.
DaCe has been tested with Intel FPGA SDK for OpenCL Pro edition v18.1 and v19.1, targeting Arria 10 and Stratix 10 devices, and Xilinx Vitis HLS v2020.x, v2021.x targeting u250 and u280 devices.


**Distributed Computing**: If using multiple nodes, MPI has to be installed and available.

Expand Down Expand Up @@ -136,6 +138,12 @@ Common issues with the DaCe Python module
* **Bug in DaCe**: If you suspect an issue happens within DaCe, see :ref:`debugging` for ways to pinpoint the source
of the issue.

* **Intel FPGA libraries not found**: when targeting Intel FPGAs, the compilation process may fail due to missing OpenCL headers (CMake returns
a ``Could NOT find IntelFPGAOpenCL`` error). This is usually the case when Intel OpenCL compiler does not return the right path to OpenCL host headers.
DaCe relies on ``hlslib`` for compiling FPGA programs, which in turns relies on Intel's compiler to derive the right include path. Please verify that
the include path returned by the Intel compiler (using the ``aocl compile-config`` command) points to a directory that actually contains the OpenCL headers (namely ``cl.hpp`` and
``cl2.hpp`` files). If this is not the case, please locate them under the Intel Quartus installation folder, and symlink (or copy) them in the ``aocl`` returned path.

.. _qa_vscode:

Common issues with the Visual Studio Code extension
Expand Down
Loading