The Alveo U30 media accelerator card provides the industry’s highest channel density, lowest cost per channel, and lowest power consumption for live video streaming workloads. When coupled with Xilinx’s Video Software Developers Kit (SDK) the solution provides a production-ready platform for live streaming providers to benefit from Xilinx hardware acceleration with standards-based FFmpeg, GStreamer and APIs.
The U30 is available for live video transcoding in the cloud over Amazon EC2 VT1 instances or as a preconfigured appliance for on-premises deployment from our partners.
Built for Performance & Efficiency
Adaptable with Ease of Integration
Accessible - Cloud <-> On-Premises Mobility
|Form Factor, Passive Form Factor, Active||Half Height, ½ Length|
|Logic Resources2||Look-Up Tables||460K|
|Video support||Video Codec Unit (VCU)||Hardened|
|Codec||H.264 and H.265|
|Video Transcodes per Card||2 x 4kp60, 8 x 1080p60,16 x 1080p30, 32 x 720p30|
|DRAM Memory||DDR Format||2 x 4GB 72b DDR4|
|DDR Total Capacity||8GB|
|DDR Max Data Rate||2400MT/s|
|DDR Total Bandwidth||38GB/s|
|Internal SRAM||Total Capacity||11MB|
|Interfaces||PCI Express®||Gen3 x8, 2 x Gen3 x43|
|Power and Thermal||Thermal Cooling||Passive|
|Typical Power||18-25W typical4|
|Tool Support||Vivado® Design Suite||No|
|FFmpeg with Xilinx Video Plugins||Yes|
|Vitis™ Developer Environment||No|
|Solutions||Solutions & Libraries||Video Transcoding Solutions|
Accelerate your streaming workflows with our turnkey software developers kit (SDK). Access the tools, APIs, and high-level FFmpeg frameworks, GStreamer, or your own proprietary framework via C/C++ APIs and get up and running in just one day.
The U30 Video SDK is a complete software stack allowing users to seamlessly leverage the hardware accelerated features of Alveo U30 cards for on-premises or cloud deployments. It includes the following elements:
Full documentation including installation guide, tutorials and demo videos available on GitHub
Amazon EC2 VT1 instances are designed to accelerate real time video transcoding and deliver low cost transcoding for live video streams.
They can deliver up to 30% lower cost per stream compared to Amazon EC2 G4dn GPU-based instances and up to 60% lower cost per stream compared to Amazon EC2 C5 CPU-based instances for transcoding live video streams.
Gbps of high speed networking and storage
Lower cost vs GPU
Simultaneous 1080p60 streams in real time
Support for up to 4K UHD at 60fps