AMD Advanced Media Framework

Optimal access to AMD GPUs for multimedia processing

The Advanced Media Framework SDK provides developers with optimal access to AMD GPUs for multimedia processing. 

Download the latest version - v1.4.35.0

This release adds the following features:

  • Added multi-HW instance encoder mode.
  • Updated FFmpeg to 7.0.
  • Added split frame encoding.

Benefits

The AMF SDK may be used in developing wireless display, remote desktop, video editing, transcode and playback applications. Both cloud gaming and remote gaming applications are also possible with the AMF SDK. Specifically, developers can use the AMF SDK for accessing AMD media accelerators for video encoding, decoding, color space conversion, pre-analysis, pre-processing, display capture and high quality scaling.

The AMF SDK allows optimization of application performance by utilizing CPU, GPU compute shaders and hardware accelerators for media processing. These optimizations are applicable to a wide range of applications such as gaming or content creation. Programming of AMD Video Engines (UVD and VCE/VCN blocks) is also an important part of the functionality that AMF provides to developers.

Requirements

  • Windows®
    • Windows® 7 (SP1 with the Platform Update) (AMF v1.4.18.0 and older).
    • Windows® 8.1 (AMF v1.4.0.0 and older).
    • Windows® 10, or Windows® 11.
    • Windows Subsystem for Linux (DX12 Decoder and Converter Only).
    • Visual Studio® 2019.
  • Linux
    • RHEL 9.4 / 8.10
    • Ubuntu 24.04.1 / 22.04.5
    • SLED/SLES 15 SP6
  • Driver and AMF component installation instructions are available on the Wiki page.
    • The following table contains the driver versions in which the Linux pro driver started including the AMF runtime, otherwise, the AMF runtime is optional and has to be installed separately.
    • All supported distros include the AMF runtime starting driver version 20.40.
OS AMF included starting version
SLE 15 18.40
Ubuntu 20.04.0 20.20
*All supported distros 20.40
  • AMF SDK is backward compatible with all previous driver versions.
    • Version 1.4.35: AMD Radeon Software Adrenalin Edition 24.9.1 (24.20.11.01) or newer. Added multi-HW instance encoder mode. Updated FFmpeg to 7.0. Added split frame encoding.
    • Version 1.4.34: AMD Radeon Software Adrenalin Edition 24.6.1 (24.10.21.01) or newer. Added Linux support for DVR sample. New HEVC header insertion mode. Stable support for RADV drivers for AMF on Linux in VideoConverter/HQScaler/VideoEncoder and experimental for decoder. Added DX11 support for FRC component.
    • Version 1.4.33: AMD Radeon Software Adrenalin Edition 24.1.1 (23.40.02) or newer. Added native DX12 support for encoding and PreAnalysis. Switched to public Vulkan Khronos extensions for decoder. AMF on Linux can now be used with AMD Pro Vulkan, and experimentally with RADV drivers.
    • Version 1.4.32: AMD Radeon Software Adrenalin Edition 23.12.1 (23.30.13.01) or newer. Added new FRC component and API doc. Added new VQEnhancer doc. Improved presenters. Updated FFmpeg to 6.0.
    • Version 1.4.30: AMD Radeon Software Adrenalin Edition 23.5.2 (23.10.01.45) or newer. Added AMF wrappers for AVC / HEVC / AV1 FFmpeg software encoders, frame in -> slice / tile output support for AVC / HEVC / AV1 encoders, and multi-monitor support for DVR. Updated FFmpeg to 5.1.2.
    • Version 1.4.29: AMD Radeon Software Adrenalin Edition 23.1.2 (22.40.01.34) or newer. Added Smart Access Video for AVC / HEVC / AV1. New options for VQEnhancer and AV1 encoder components. Switched to Markdown based API docs which enable easier open source developer contributions.
    • Version 1.4.28: AMD Radeon Software Adrenalin Edition 22.12.1 (22.40.00.24) or newer. Added AV1 encoding support and 12-bit AV1 decoding. New VQEnhancer component.  New AVC/HEVC encoder rate control methods.
    • Version 1.4.26: AMD Radeon Software Adrenalin Edition 22.7.1 (22.20.15.01) or newer. Added new PAQ, TAQ, and high motion quality boost modes for PreAnalysis. New HQScaler sharpness, low latency decoder and temporal SVC encoder options.
    • Version 1.4.24: AMD Radeon Software Adrenalin Edition 22.3.1 (21.50.11.03) or newer. Added new AMD Direct Capture mode, new HQscaler feature(Bilinear/Bicubic/FSR), new Vulkan HEVC encoder on Navi family, improvements on H264 Vulkan encoding.
    • Version 1.4.23: AMD Radeon Software Adrenalin Edition 21.12.1 (21.40.11.03) or newer. Added new Auto LTR encoder mode, additional encoder usage presets and encoder statistics/feedback.
    • Version 1.4.21: AMD Radeon Software Adrenalin Edition 21.10.1 (21.30.25.01) or newer. Added PSNR/SSIM score feedback, new QVBR rate control mode and LTR mode for encoders, added HDR support for HEVC encoder and color converter, new EncoderLatency sample app.
    • Version 1.4.18: AMD Radeon Software Adrenalin Edition 20.11.2 or newer. Added Pre-Encode filter within Pre-Processing component in 1.4.18.
    • Version 1.4.9 or later requires Vulkan SDK for some samples: https://vulkan.lunarg.com/ and AMD Radeon Software Adrenalin Edition 18.8.1 (18.30.01.01) or newer. This version supports Linux (see amd.com for driver support).
    • Version 1.4.4 or later requires OCL_SDK_Light: https://github.com/GPUOpen-LibrariesAndSDKs/OCL-SDK/releases. Previous versions of AMF require the AMD APP SDK (Version 3.0 or later), Windows 10 SDK (Version 10586). This version requires AMD Radeon Software Crimson Edition 17.7.2 (17.30.1041) or newer.
    • Version 1.4: AMD Radeon Software Crimson Edition 17.1.1 (16.50.2611) or newer.
    • Version 1.3: AMD Radeon Software Crimson Edition 16.7.3 (16.30.2311) or newer.

Version history

  • Added multi-HW instance encoder mode.
  • Updated FFmpeg to 7.0.
  • Added split frame encoding.
  • Added Linux support for DVR sample.
  • New HEVC header insertion mode.
  • Stable support for RADV drivers for AMF on Linux in VideoConverter/HQScaler/VideoEncoder and experimental for decoder.
  • Added DX11 support for FRC component.
  • Added native DX12 support for encoding and PreAnalysis
  • Vulkan encoder became independent from Vulkan driver
  • Switched to public Vulkan Khronos extensions for decoder
  • AMF on Linux can now be used with AMD Pro Vulkan, and experimentally with RADV drivers
  • Improvements to sample presenters
  • Added new FRC component and API doc
  • Added new VQEnhancer API doc
  • Improvements and cleanup in sample presenters
  • Updated FFmpeg to 6.0
  • Added AMF wrappers for AVC / HEVC / AV1 FFmpeg software encoders
  • Added frame in, slice / tile output support for AVC / HEVC / AV1 encoders
  • Added multi-monitor support for DVR app
  • Added decoder relocation between VCN instances and new capability query for SmartAccess Video
  • New AV1 usages and encoder feedback
  • Improvements and cleanup to sample apps
  • Updated FFmpeg to 5.1.2
  • Added Smart Access Video for decoder and AVC / HEVC / AV1 encoders
  • New options for VQEnhancer and AV1 encoder components
  • Switched to Markdown based API docs which enable easier open source developer contributions
  • AV1 encoding
  • 12-bit AV1 decoding
  • New VQEnhancer component
  • New AVC / HEVC encoder rate control methods
  • New perceptual (PAQ) and temporal (TAQ) adaptive quantization modes for PreAnalysis
  • New high motion quality boost mode for PreAnalysis
  • New sharpness option for HQScaler
  • New low latency decoding option
  • New temporal SVC encoding options
  • New AMD Direct Capture component and updated DVR sample app
  • New HQ Scaler ( Bilinear / Bicubic / FSR )
  • Vulkan support for PreAnalysis component
  • Vulkan support for HEVC Encoder ( Navi+ )
  • Support for temporal layer query for AVC and HEVC Encoder
  • Support for consecutive B-frames and adaptive miniGOP for AVC Encoder
  • Improved H264 Encoding on Vulkan ( both Windows & Linux Navi+ )
  • New Auto LTR encoder mode.
  • FFmpeg updated to version 4.3.2.
  • Additional statistics and feedback in encoder.
  • New encoder usage presets.
  • Main 10 HEVC encoder profile support.
  • Support for encoder instance selection.
  • HDR color conversion in the encoder.
  • New QVBR rate control mode and LTR encoder mode.
  • Extended statistic and quality metrics in the encoder.
  • Frame rate parameter became dynamic and HDR color conversion was updated.
  • Updated samples with encoder latency functionality.
  • AMD Radeon™ RX 6000 series graphic cards.
  • Decode and color conversion support within DirectX® 12 pipeline.
  • Decode of VP9 and AV1 files on supported hardware.
  • Encode features to support importance maps, picture management, output timeout, statistics and throughput ceilings.
  • Pre-Encode filter within Encode Pre-Processing component.
  • Updated samples with Pre-Encode filter and frame injection functionality.
  • Encode pre-analysis component to improve the perceived quality of encoded video.

Requires AMD Radeon™ Software Crimson 16.7.3 (16.30.2311) or later

Our other SDKs

Anti-Lag 2 SDK

AMD Radeon™ Anti-Lag 2 reduces the system latency by applying frame alignment between the CPU and GPU jobs.

AMD Capsaicin Framework

Capsaicin is a Direct3D12 framework for real-time graphics research which implements the GI-1.0 technique and a reference path-tracer.

AMD Render Pipeline Shaders (RPS) SDK

The Render Pipeline Shaders (RPS) SDK provides a framework for graphics engines to use Render Graphs with explicit APIs.

AMD Device Library eXtra

ADLX is a modern library designed to access features and functionality of AMD systems such as Display, 3D graphics, Performance Monitoring, GPU Tuning, and more.

AMD Brotli-G SDK

Brotli-G is an open-source compression/decompression standard for digital assets (based on Brotli) that is compatible with GPU hardware.

AMD HIP Ray Tracing

HIP RT is a ray tracing library for HIP, making it easy to write ray tracing applications in HIP.

AMD Orochi

Orochi is a library which loads HIP and CUDA® APIs dynamically, allowing the user to switch APIs at runtime.

AMD Radeon ProRender Developer Suite

AMD Radeon™ ProRender is our fast, easy, and incredible physically-based rendering engine built on industry standards that enables accelerated rendering on virtually any GPU, any CPU, and any OS in over a dozen leading digital content creation and CAD applications.

AMD Radeon ML

Radeon™ Machine Learning (Radeon™ ML or RML) is an AMD SDK for high-performance deep learning inference on GPUs.

AMD Radeon Image Filter

Harness the power of machine learning to enhance images with denoising, enabling your application to produce high quality images in a fraction of the time traditional denoising filters take.

AMD GPUOpen Direct3D12 Memory Allocator (D3D12MA)

The D3D12 Memory Allocator (D3D12MA) is a C++ library that provides a simple and easy-to-integrate API to help you allocate memory for DirectX®12 buffers and textures.

ADL

The AMD Display Library (ADL) SDK is designed to access display driver functionality for AMD Radeon™ and AMD FirePro™ graphics cards.