AMD VCE/VCN Hardware Accelerated Encoding Decoding

Have you ever heard of Quick Sync Video or NVENC? Maybe CUDA sounds familiar. All of the three are hardware acceleration technologies developed by Intel and Nvidia, respectively. AMD also has its own unique hardware-accelerated video encoding/decoding technique. That was VCE, now evolved into VCN. On this page, you will know what VCE/VCN is, and what AMD GPUs/APUs support VCE.

What Is AMD VCE / VCN?

Generally speaking, VCE is a kind of hardware-accelerated video encoding mode in AMD. Surely there is a hardware-accelerated video decoder as well — UVD (Unified Video Decoder). AMD integrates a Multimedia Accelerator in its GPUs and APUs, inside which ASIC[1] — VCE implements H.264 / MPEG-4 AVC video encoding only.

[1] ASIC — Application-specific integrated circuit

Other Names of VCE

1. VCE is the abbreviation of Video Coding Engine.
2. There are some other full names: Video Coding Engine, Video Compression Engine, or Video Codec Engine.
3. VCE is also known as AMF — Advanced Media Framework.

The Development of AMD VCE

There are four major updates of AMD Video Coding Engine technology. Each version was applied to different APUs and GPUs. VCE supports encoding H.264 only until Version 3.0. HEVC/H.265 was added.

How Does AMD VCE Accelerated Video Encoding?

There are two hardware-accelerated video encoding modes when using AMD VCE — fixed mode and hybrid mode. Both have their merits and cons. See the following details.

VCE Full-Fixed Mode

1. The whole video encoding process, including the intra prediction, motion estimation, and entropy encode is done through the VCE circuit in ASIC.
2. Full-fixed mode is accessed through OpenMAX IL API.
3. Merits: Fast video encoding speed.
4. Cons: For simply editing only with a low compression ratio. You may get a huge-size file.

VCE Hybrid Mode

1. Actions like intra prediction, motion estimation, quantization, and forward transform are done through a 3D engine, while the final entropy encode uses VCE.
2. Hybrid mode is accessed through AMD’s Accelerated Parallel Programming SDK and OpenCL.
3. Merits: High compression ratio and good for complicated video editing.
4. Cons: Slow encoding speed.

AMD Graphics Cards That Support VCE / VCN

Check all the graphics cards made by AMD that support VCE hardware acceleration in the table below. Besides, there is information about the maximum resolution supported in VCE and video decoder.

Codecs That Can Use AMD VCN 1.0 Hardware Acceleration

As we mentioned before, VCE has been replaced by VNC as of Raven Ridge in 2017. What is the difference? VCE is hardware acceleration utilized in video encoding only, while VNC is for both hardware-accelerated encoding and decoding. Let’s see what codecs can be decoded and encoded using VNC.

AMD APUs That Support VCE / VCN

VCE and VCN hardware acceleration is not only used in AMD graphics cards but also in APUs[2]. Not sure whether your AMD computers with APU can decode and encode video in VCE / VCN? Check the following lists.

[2] APU — Accelerated Processing Unit

VideoProc Converter — Makes All VCE, NVENC, QSV Acceleration Accessible!

VCE seems to be not as well known as Nvidia NVENC and Intel QSV. Why is that? The biggest reason might be that few programs support AMD VCE while both Nvidia CUDA/NVENC and Intel QSV are widely compatible.

VideoProc Converter is a video conversion tool, which gives support on full use of AMD VCE/VNC, Nvidia CUDA/NVENC, and Intel QSV. Such hardware acceleration is particularly important when decoding, editing, and encoding UHD files, like 4K video.