MOG now offers our customers the ability to accelerate media encoding and decoding operations using GPU capabilities.
Technically speaking, hardware acceleration is the process by which an app will use other hardware components on your system to perform certain tasks to work more efficiently.
Here are some facts about the benefit of using Hardware Acceleration:
When it comes to decoding, Windows 10 includes support for native GPU accelerated video decode using DXVA 2 (https://docs.microsoft.com/en-us/windows/win32/medfound/about-dxva-2-0). This API offers an out-of-box cross GPU implementation for an easier integration without having to explicitly use a driver or vendor solution. In the background, the correct and highest performance module is selected, or a CPU fallback if the profile isn’t supported.
Depending on the hardware of the server where MAM4PRO is installed, the following codecs can use hardware decoding:
Some configurations can be enabled, to choose in which component the decoding should be accelerated: Player, Ingest, or both. By default, both options are enabled.
For the encoding, MAM4PRO has now integrated NVENC (Nvidia NVENC) and QSV (Intel Quick Sync Video) features to perform video encoding, offloading this compute-intensive task from the CPU.
This can be enabled when configuring a workflow profile, by setting the Encoder Family to one of these options. In this case, unlike decoding, if the profile is not supported, there is no CPU fallback.
The following codecs can use hardware encoding:
In the following diagram, it is represented how and in which modules this feature is implemented.