yugang-amd
46 Zeilen
2.2 KiB
ReStructuredText
46 Zeilen
2.2 KiB
ReStructuredText
.. meta::
|
|
:description: rocSHMEM intra-kernel networking runtime for AMD GPUs on the ROCm platform.
|
|
:keywords: rocSHMEM, API, ROCm, documentation, HIP, Networking, Communication
|
|
|
|
.. _rocshmem-introduction:
|
|
|
|
---------------------------
|
|
What is rocSHMEM?
|
|
---------------------------
|
|
|
|
The ROCm OpenSHMEM (rocSHMEM) is an intra-kernel networking library that provides GPU-centric networking through an OpenSHMEM-like interface. It simplifies application code complexity and enables finer communication and computation overlap than traditional host-driven networking. rocSHMEM uses a single symmetric heap allocated on GPU memories.
|
|
|
|
The rocSHMEM programming model
|
|
-------------------------------
|
|
|
|
Defining how OpenSHMEM applications interact with GPUs remains an
|
|
ongoing active discussion within the OpenSHMEM community, and the OpenSHMEM
|
|
specification has yet to coalesce on this topic.
|
|
rocSHMEM extends beyond the OpenSHMEM specification to add semantics that
|
|
support GPU kernel communication while maintaining close resemblance to
|
|
the original OpenSHMEM specification semantics.
|
|
|
|
Applications using :doc:`HIP <hip:index>` can interface with rocSHMEM.
|
|
Using the HIP programming model,
|
|
rocSHMEM provides ``__host__`` APIs for host code,
|
|
and ``__device__`` APIs for GPU kernels.
|
|
Device APIs without special suffixes or infixes , for example, ``_wg`` or ``_wave``,
|
|
must be called by a single thread.
|
|
GPU specific ``_wg`` and ``_wave`` APIs are designed to be called by multiple GPU threads
|
|
and will block until the calling scope completes.
|
|
These APIs can be called in divergent code paths, but this is not recommended.
|
|
|
|
Wavefront APIs
|
|
==============
|
|
Wavefront APIs are those with the ``_wave`` suffix.
|
|
The parameters in which these routines are called must be
|
|
the same for every thread in the wavefront.
|
|
The behavior is undefined if any thread calls these routines with different parameters. These APIs will block until the calling wavefront is complete.
|
|
|
|
Workgroup APIs
|
|
==============
|
|
The workgroup APIs have the ``_wg`` suffix or ``_wg_`` infix.
|
|
The parameters in which these routines are called must be
|
|
the same for every thread in the workgroup.
|
|
The behavior is undefined if any thread calls these routines with different parameters. These APIs will block until the calling workgroup is complete.
|