AMD Platforms for Pro AV & Broadcast
All AMD devices and platforms are usable across Pro AV and Broadcast, but there are some specific devices of interest when it comes to capabilities ideally suited to the latest UHD applications. The Cost Optimised Portfolio is a range of devices suited to cost and power-sensitive Pro AV and Broadcast systems whilst still capable of UHD interfacing and processing. The Zynq UltraScale+ MPSoC EV family includes an integrated H.264/H.265 codec for streaming applications and is available on the Kria SOMs to speed development and deployment. Virtex UltraScale+ HBM and 58G FPGAs and Versal adaptive SoCs are ideally suited to high-performance 8K or multichannel 4K processing and up to 100GE networking. And Alveo accelerator cards provide a new way to deploy server-based media functions such as transcoding and machine learning either on-premise or in the cloud.
COST OPTIMIZED PORTFOLIO
Our Cost Optimized Portfolio can efficiently implement broadcast-quality 4K60 UHD video pipelines for space-constrained applications, such as mini-converters, cameras and KVM or other high-volume cost-sensitive applications. Whether using AMD FPGAs or SoCs, massively parallel DSP ensures scalable real-time video processing for multiple AV channels in a single device, and high-performance connectivity enables bridging between SDI, HDMI, DisplayPort, and 1Gb/10Gb Ethernet for live 4K AV-over-IP streaming.
ZYNQ ULTRASCALE+ MPSOC WITH VIDEO CODEC UNIT
The Zynq UltraScale+ EV MPSoC family is unique in the industry. It includes a hardened H.264/H.265 video codec unit (VCU) supporting up to 4K60 4:2:2 10-bit simultaneous encode and decode (and therefore transcode), or up to 32 streams of lower resolution video. It combines excellent video quality with the industry’s lowest latency for a professional H.264/H.265 codec, as well as balancing with low power implementation. The codec is full featured with hundreds of options for different streaming use cases. We use GStreamer and OpenMAX software frameworks for ease-of-use as well as offering control software for fine tuning parameters. Plus, there is programmable logic, high-speed transceivers and ARM processors to integrate AV processing pipelines in the same device, making it the ultimate streaming platform.
- Learn more about the Zynq UltraScale+ MPSoC family >
- Single-Chip Streaming Part 1 - Jump-start for Complex Multimedia Streaming Designs >
- Single-Chip Streaming Part 2 - Adaptable High Dynamic Range Streaming >
- Single Chip Streaming Part 3 - 4:4:4 Full Chroma Streaming for Video with Desktop Content >
- Single-Chip Streaming Part 4 - Configurable Low Latency Streaming >
Zynq UltraScale+ MPSoC Video Codec Resources
Documentation
AMD Videos
More Resources
KINTEX ULTRASCALE+ FPGA
The Kintex UltraScale+ FPGA family has proven to be ideal for many cost-effective broadcast and pro AV systems that require 25Gb Ethernet networking. Supporting up to 33Gbps, the transceiver performance is unrivalled at this price point, and connectivity to 25GE, UHD-SDI, HDMI and DisplayPort, as well as PCIe® Gen4, means that it is an ideal video and audio processing platform needing the best price/performance/watt balance.
VIRTEX ULTRASCALE+ FPGA
To help drive the next wave of Ethernet deployment, AMD has integrated 58Gbps transceivers into the Virtex UltraScale+ FPGA family. With superior signal integrity, Virtex UltraScale+ FPGAs offer support for 50/100/200/400G optics and protocols via enhanced continuous adaptive RX equalization and built-in KR-FEC. Furthermore, Virtex UltraScale+ HBM FPGAs provide up to 16GB of high-bandwidth memory (HBM) Gen2 integrated in-package for 460GB/s of memory bandwidth, delivering 20X more bandwidth than a DDR4 DIMM and removing the bottlenecks associated with packet and video frame buffering.
As the broadcast industry moves to platform-based designs with greater capabilities, the integration of higher-speed Ethernet connectivity and high-bandwidth memory makes it a superb option for multichannel UHD processing and networking systems.
Versal Adaptive SoCs
The Versal adaptive SoCs offer a number of exciting new features for broadcast and high-end professional AV. The programmable network-on-chip (NoC) enables high bandwidth terabit transport of AV between processing kernels across the device. DDR4 and LPDDR4 interfaces with hardened memory controllers, and integrated high-bandwidth memory (HBM) offer the highest performance buffers for IP packets or UHD video frames. Tuned and optimized transceivers support the latest UHD-SDI, HDMI and DisplayPort standards, PCIe Gen4/Gen5 and from 10GE to 400GE, with a built-in multi-rate Ethernet MAC to support ST 2110 AV-over-IP.
This is all in addition to multicore scalar engines for OS, adaptable engines for real-time video processing and low latency compression, and intelligent engines for DSP and machine learning acceleration.
KRIA SOM
To speed development and production deployment in edge applications, AMD has introduced Kria System-on-Modules (SOMs). The first in the range includes the AMD video codec unit (VCU) which handles 1x 4Kp60 stream or up to 32 lower resolution streams, as well as providing various sensor interfaces, plus HDMI and DisplayPort connectivity. Accelerated applications are available in the Kria App Store, enabling developers to program and differentiate their designs at the software level, without requiring FPGA programming experience. Paired with the KV260 Vision AI Starter Kit, production-ready accelerated applications can be up and running in under 1 hour, allowing developers to focus on differentiation and shortening the runway to volume deployment.
ALVEO ACCELERATOR CARDS
AMD Alveo PCIe cards are aimed at server-class acceleration both on-prem and in the cloud. With abstraction away from the underlying hardware, it’s possible to run complex transcoding with just a few lines of FFmpeg instructions or perform high-quality or high-density encoding for many live streaming applications. The use of open frameworks and easily accessible APIs means that Alveo cards can be used in a variety of stages through the media workflow, from post-production and analytics, in the network with Alveo Smart NICs, through to distribution. AMD has also introduced the MA35D accelerator card to provide Alveo acceleration in an appliance that can also be used in high-density, low-latency and interactive OTT streaming applications.
AMD Applications
From AV-over-IP distribution and KVM, to collaboration, live streaming and LED walls, Pro AV systems require high quality, low-latency audio and video processing, transportation and presentation. And from acquisition to contribution, through production and post-production, to distribution and consumption, AMD adaptable platforms are at the heart of all applications used throughout cinema and broadcast.
AV Technologies
Most multimedia systems use a combination of AV connectivity (both baseband and IP transport), audio and video processing, and compression. AMD and our Alliance Program Partners offer a huge portfolio of IP cores, reference designs and subsystems to get you faster to market.
