Your Browser is Out of Date

Nytro.ai uses technology that works best in other browsers.
For a full experience use one of the browsers below
Computer Vision Workflow Enhancement – Image Restoration with ProHawk
Overview Solution overview Partner overview Architecture Sizing and scaling guidance
Sizing the solution
Scaling the solution
Use cases Conclusions and References

Scaling the solution

Home > Workload Solutions > Computer Vision > White Papers > Computer Vision Workflow Enhancement – Image Restoration with ProHawk > Scaling the solution

Scaling the solution

  • The solution has some challenges integrating with the CV platform. The ProHawk server has competing requirements for both CPU and GPU. If a system is already fully occupied performing CV workloads, it means that there are no excess CPU and GPU resources available. In this case, the options are:

    1. Reduce the number of cameras running CV analytics to free up capacity for ProHawk.
    2. Add additional Hardware to facilitate ProHawk.

    The most likely scenario is that additional hardware is added to support ProHawk. The standard server size for the CV platform is as follows:

    CPU
    2 x Intel(R) Xeon(R) Gold 6354 CPU @ 3.00 GHz (72 vCPU)
    Memory
    512 GB
    GPU
    2 x A16 64GB Nvidia GPU

    For every VxRail node we add we have the capacity to create 8 ProHawk server VMs which ensure support for 16 real-time ProHawk streams.

    In the case that only the Milestone XProtect Smart Client plug-in is used then eight Smart Clients can be configured with ProHawk restoration capabilities. Each Smart Client can show multiple restored streams at once.

    It is also possible to use ProHawk Vision servers for real-time processing and Milestone XProtect Smart Client plugins for the restoration of historic data. The exact blend is based on the users requirements.