Frequently Asked Question

NVR: How to enable GPU acceleration on Surveillance Station Client
Last Updated 7 years ago

Overview

GPU acceleration is supported on the Windows version of Surveillance Station Client 1.1 and above. By leveraging GPU to play videos, CPU loading can be efficiently reduced. At the present time, most of the Intel Core i processors are equipped with integrated GPU, therefore, by simply enabling GPU acceleration and configuring a suitable setting, streaming display will be smoother and the number of supported decoding channels will also be increased.

1. Before You Start

Before you start, please make sure the following requirements are met:

  1. System requirements:
    • Supports Windows 7 and above operating systems.
    • GPU support for DirectX Video Acceleration (DXVA) 2.0 is required.
  2. Supported GPU models:
    • This function is not restricted to specific GPU brands or models, and should work with any GPU models supporting DXVA 2.0 and the video codec you are using.
    • The decoding performance of integrated GPUs will not be lower than that of discrete GPUs according to the actual test results. We suggest trying the integrated GPUs first.

2. Configuration Introduction

  • To configure GPU acceleration, go to Options > GPU Acceleration.
  • The current GPU model will be displayed in the page description under GPU Acceleration, please refer to the image below.

    Note:
    • Surveillance Station Client will use only one GPU card to decode videos even if multiple GPU cards are installed.

  • Select Intelligent adjustment to automatically distribute GPU accelerated video streaming loads according to the decoding capabilities of the GPU to optimize system performance.
  • Select Manual adjustment to choose a specific number of streams and the minimum resolution of the streams required to perform GPU decoding. If the actual number of live view streams exceeds the specified number, additional streams will be decoded using CPU.
  • Tick the Label the video streams that are using GPU decoding on the screen checkbox to identify streams that are decoded through GPU in Live View. A GPU icon will be displayed on the upper-right corner of the video streams that are using GPU decoding, as shown below:

3. Currently Supported Video Formats and Surveillance Station Applications

  • Only the H.264 video format supports GPU acceleration. Other video formats will still use CPU decoding.
  • Live View, Preview, Timeline, and recording playback all support GPU decoding.
  • Note:
    • Advanced video processing functions such as fisheye dewarping, and streams in Live View Analytics still require CPU decoding.

4. Performance Optimization Suggestion

If video streams in Live View cannot be displayed smoothly, you can make adjustments following the steps below:

  1. Go to Options > GPU Acceleration and select the Intelligent adjustment option first for smoother streaming display.
    • This option satisfies most user requirements under usual usage conditions.
  2. If video streaming cannot be displayed smoothly when selecting the Intelligent adjustment option, you can select the Manual adjustment option to optimize performance. Please refer to the steps below to determine the number for maximum channel count:
    • Start with a lower channel count to test the streaming quality (e.g. open one channel in Live View, and set the maximum channel count to one). If the streaming quality is smooth, then increase the channel count until live view streams start to lag.
    • For example, if "x" is the maximum number of channels in which live view streams can still be displayed smoothly, then you can enter "x" in the maximum channel count field. "x" channels in Live View will then be the maximum number of channels that will use GPU decoding, and when exceeding the maximum number of channels, then additional channels will use CPU decoding.
    Note:
    • When determining the ideal "maximum number of channels", it may be difficult to determine if image latency is caused by CPU or GPU performance if Live View includes streams that are decoded by CPU and streams decoded by GPU. Therefore, to avoid confusion, it is recommended to ensure all streams use GPU decoding to create a simplified testing environment.

5. Performance Reference

You can refer to the data listed in this article to pick your hardware. The Manual adjustment option is used during the test

Please Wait!

Please wait... it will take a second!