File: README.md

package info (click to toggle)
libxsmm 1.17-4
  • links: PTS, VCS
  • area: main
  • in suites: sid, trixie
  • size: 14,976 kB
  • sloc: ansic: 119,587; cpp: 27,680; fortran: 9,179; sh: 5,765; makefile: 5,040; pascal: 2,312; python: 1,812; f90: 1,773
file content (25 lines) | stat: -rw-r--r-- 2,131 bytes parent folder | download | duplicates (2)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
# Dispatch<a name="dispatch-microbenchmark"></a>

## Microbenchmark

This code sample benchmarks the performance of (1)&#160;the dispatch mechanism, and (2)&#160;the time needed to JIT-generate code for the first time. Both mechanisms are relevant when replacing GEMM calls (see [Call Wrapper](https://libxsmm.readthedocs.io/libxsmm_mm/#call-wrapper) section of the reference documentation), or in any case of calling LIBXSMM's native [GEMM functionality](https://libxsmm.readthedocs.io/libxsmm_mm/).

**Command Line Interface (CLI)**

* Optionally takes the number of dispatches/code-generations (default:&#160;10000).
* Optionally takes the number of threads (default:&#160;1).

**Measurements (Benchmark)**

* Duration of an empty function call (serves as a reference timing).
* Duration to find an already generated kernel (cached/non-cached).
* Duration to JIT-generate a GEMM kernel.

In case of a multi-threaded benchmark, the timings represent a highly contended request (worst case). For thread-scaling, it can be observed that read-only accesses (code dispatch) stay roughly with a constant duration whereas write-accesses (code generation) are serialized and hence the duration scales linearly with the number of threads.

The [Fortran example](https://github.com/hfp/libxsmm/blob/master/samples/utilities/dispatch/dispatch.f) (`dispatch.f`) could use `libxsmm_dmmdispatch` (or similar) like the C code (`dispatch.c`) but intentionally shows the lower-level dispatch interface `libxsmm_xmmdispatch` and also omits using the LIBXSMM module. Not using the module confirms: the same task can be achieved by relying only on FORTRAN&#160;77 language level.

## User-Data Dispatch

Further, another [Fortran example](https://github.com/hfp/libxsmm/blob/master/samples/utilities/dispatch/dispatch_udt.f) about [user-data dispatch](https://libxsmm.readthedocs.io/libxsmm_aux/#user-data-dispatch) is not exactly a benchmark. Dispatching user-data containing multiple kernels can obviously save multiple singular dispatches. The C interface for dispatching user-data is designed to follow the same flow as the Fortran interface.