LikwidMarkerAPIPitfalls.md

@@ -102,7 +102,9 @@ This way, the benchmark can execute the array copy without influences from the M
 # The measured times for multiple threads vary although all perform the same operation
 
 In some cases, you might see measurements like this:
-ADD STREAM TRIAD WITHOUT REGISTER
+
+![Measurement of STREAM triad over 9 threads without LIKWID_MARKER_REGISTER](https://raw.githubusercontent.com/wiki/RRZE-HPC/likwid/images/LikwidMarkerAPIPitfalls/stream_compare_band_no_register.png)
+
 
 Let's look at one output (just an excerpt):
 ```