1,433
edits
AM5728 Multimedia Performance Testbench: Difference between revisions
AM5728 Multimedia Performance Testbench (view source)
Revision as of 20:23, 7 June 2016
, 7 June 2016→AAC audio encode
| Line 63: | Line 63: | ||
=== <span style="color:#0931C6"> | === <span style="color:#0931C6">Buffer processing time</span><br> === | ||
'''''Test pipeline:''''' | '''''Test pipeline:''''' | ||
<pre style="background:#d6e4f1"> | <pre style="background:#d6e4f1"> | ||
GST_TRACER_PLUGINS=" | GST_TRACER_PLUGINS="proctime" gst-launch-1.0 filesrc location=/am5728-gst-tests/audio-samples/audio_sample.raw ! audioparse ! faac ! fakesink -e | ||
</pre> | </pre> | ||
'''''Obtained Results:''''' | '''''Obtained Results:''''' | ||
[[Image:AM572x-testbench-AAC- | [[Image:AM572x-testbench-AAC-buffertime.png|center|700px|AM572x-testbench-AAC-buffertime.png]]<br> | ||
In the chart above, it can be seen that when using hardware acceleration (NEON&VFPv4 extension), | In the chart above, it can be seen in a general way that when using hardware acceleration (NEON&VFPv4 extension), a lees buffer processing time is achieved. The average difference per buffer is 17.47 us when hardware acceleration (NEON&VFPv4 extension) is used. | ||
=== <span style="color:#0931C6"> | === <span style="color:#0931C6">CPU load % per core</span><br> === | ||
'''''Test pipeline:''''' | '''''Test pipeline:''''' | ||
<pre style="background:#d6e4f1"> | <pre style="background:#d6e4f1"> | ||
GST_TRACER_PLUGINS=" | GST_TRACER_PLUGINS="cpuusage" gst-launch-1.0 filesrc location=/am5728-gst-tests/audio-samples/audio_sample.raw num-buffers=1000 ! audioparse ! faac ! fakesink sync=true -e | ||
</pre> | </pre> | ||
'''''Obtained Results:''''' | '''''Obtained Results:''''' | ||
[[Image:AM572x-testbench-AAC- | [[Image:AM572x-testbench-AAC-cpuload.png|center|700px|AM572x-testbench-AAC-cpuload.png]]<br> | ||
In the chart above, it can be seen | In the chart above, it can be seen that when using hardware acceleration (NEON&VFPv4 extension), no reduction is achieved in the CPU workload. The average difference between CPU_0_accel and CPU_1_unaccel is 0.5 % more load for CPU_0_accel. The average difference between CPU_1_accel and CPU_0_unaccel is 0.125 % more load for CPU_1_accel. | ||
| Line 132: | Line 132: | ||
In the chart above, it can be seen that when using hardware acceleration (NEON&VFPv4 extension), more memory bandwidth consumption by memory writings is obtained. The average difference is 206.8 MB/s for sequential reads and 79 MB/s for aleatory reads. | In the chart above, it can be seen that when using hardware acceleration (NEON&VFPv4 extension), more memory bandwidth consumption by memory writings is obtained. The average difference is 206.8 MB/s for sequential reads and 79 MB/s for aleatory reads. | ||
== <span style="color:#008080">H264 video encode</span><br> == | == <span style="color:#008080">H264 video encode</span><br> == | ||