Skip to content

Commit 7af8580

Browse files
authored
Update docs with build instructions
1 parent 24f908e commit 7af8580

File tree

1 file changed

+12
-0
lines changed

1 file changed

+12
-0
lines changed

docs/gpus.rst

Lines changed: 12 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -20,6 +20,18 @@ The HAPI Charm4py API is:
2020
For now, ``charm.hapiAddCudaCallback`` only supports numba and torch streams as input. This function inserts a callback
2121
into the stream such that when the callback is reached, the corresponding Charm4py future is set.
2222

23+
Enabling HAPI
24+
--------
25+
To build Charm4py with HAPI support, add "cuda" to the Charm build options and follow the steps to build Charm4py from source:
26+
27+
.. code-block:: shell
28+
29+
export CHARM_EXTRA_BUILD_OPTS="cuda"
30+
pip install .
31+
32+
.. warning::
33+
34+
To ensure that the underlying Charm build has Cuda enabled, remove any pre-existing builds in charm_src/charm before setting the Cuda option and running install.
2335

2436
Examples
2537
--------

0 commit comments

Comments
 (0)