1 chip decoding is a special hardware responsible for video encoding decoding, this is the basic function, there is nothing to say, look at the literal meaning you can understand. There are many encoding methods, of course, the centralized decoding method integrated in the specialized decoding chip may mainly deal with several corresponding decoding mechanisms.
2. The decoding of software vendors and the hardware decoding of chips are actually solved by processor operations from the final implementation method. It's just a different processor.
Software decoding generally relies on software to direct the cpu in the computer to calculate the corresponding instructions to decode. Of course, the cpu in the computer should be regarded as a universal processor, that is, the kind that can do anything, but nothing fine.
Because the architecture pattern of CISC was originally designed, it is a pipeline way of complex instruction sets.
This results in a relatively high probability of miscalculation, so some inefficiency, if the cpu is used to decode the video, sometimes the cpu load is very high.
However, modern cpus generally have inherited new instruction sets to correspond to operations such as video, which will improve efficiency to a certain extent.
The decoding chip receives the video encoding directly and decodes it directly by the hardware. It is generally characterized by a fixed structure for processing several corresponding video encoding formats.
The difference between him and the cpu is that he is very specialized and can only do video decoding, and basically can't do anything else. In fact, in the vernacular, the cpu is like a knife, and the decoding chip is a grid cutter.
Software decoding is the software directing the cpu knife to cut out the grid one by one. The chip decoding is to cut a bunch of grids with a knife.
Either way, it's just a matter of efficiency. Does that make sense to you?