Comparison between display settings in Citrix HDX

Hi everybody,

Last Citrix Synergy (2017) in Orlando, Benny Tritsch and Ruben Spruijt presented the following session:

SYN303: Independent Citrix experts deep dive on Remote Graphics, user experience and GPUs.
This session was among the top 10 of most popular sessions of Synergy 2017, if you didn’t see the presentation you can watch it on-demand here.

In this presentation the new Remote Display Analyzer (RDA) version 1.6 was announced which will be released soon. This version includes numerous new features like a screen capture button and a GPU add-on (available for all display protocols Remote Display Analyzer supports). More about this new RDA version soon.

Also shown in this presentation where the results of a short comparison between 3 major display settings currently available in Citrix HDX. In this blogpost I will show you how this comparison was made together with some key findings. The display settings where evaluated in real-time by playing the exact same video three times while live switching between the following settings:

  • Video codec not in use
  • Video codec for the Entire screen
  • Video codec for Actively changing regions

You can watch the comparison here:

(Test video source: Aswanth Mohan (www.youtube.com/watch?v=668nUCeBHyY))

The result of each setting is shown below:

Some key findings from this short comparison:

  • The total send frames where nearly the same
  • During the test I noticed no big difference in user experience while playing this short 720p HD video
  • Video codec not in use has the lowest impact on CPU but as you can see the bandwidth consumption is higher then the other two
  • Video codec used for the entire screen (with text optimization) gives the best result on bandwidth consumption, but this comes with a higher CPU usage
  • Video codec used for Actively changing regions results in a very good balance between both CPU and bandwidth consumption

Disclaimer:
Please note that this results come without warranty of any kind and are my own interpretation based on above test. The results may vary based on factors like workload, bandwidth and the infrastructure you are using. Please always test settings in your own environment to decide which configuration fits best in your scenario.

Advertisements