How much latency does PCoIP technology add to the remote system?

Rate this Article
No votes yet

The latency added by PCoIP technology to a remote system is related to the time to encode (compression) and decode (decompression) for the PCoIP protocol, which depends on the implementation. Total end-to-end latency is determined by OS latency and network latency. This depends on several factors:

  • length of physical media

  • switch hops

  • bandwidth

  • network congestion, etc.

PCoIP Remote Workstation Cards (Host Cards) and PCoIP Zero Client

Host cards use hardware acceleration to encode and PCoIP Zero Clients use hardware acceleration to decode the user desktop (display imaging, USB and audio). The PCoIP protocol implemented in hardware generally adds about one frame update period to the total latency, independent of image update size. Since the PCoIP Remote Workstation Card supports up to 60 frames per second, one frame time translates to 17 ms per frame. An additional period is required to synchronize the decoded image to the display timing of the monitor to prevent tearing. On average this takes about one half of a frame period or about 10 ms.

HP Anyware

The encode and decode time for the PCoIP protocol depends on the platform used on the host (encode) and client (decode). It is difficult to provide a specific latency added by the HP Anyware due to the broad range of host/client platforms used. The PCoIP Agent is configured to use 30 fps by default. As HP Anyware if completely implemented in software, the factors that influcence latency can be, but not limited to:

  • Time taken to fetch the frame buffer from the GPU (hardware or software)
  • Physical RAM speed
  • CPU utilization and clock speed
  • Encode and decode time
  • Network.