Learn how to scale your existing infrastructure to meet the demands of today’s live video streaming workloads and future-proof your streaming network in our technical eBook.
Take a deep dive to learn how to stand up an Amazon EC2 VT1 instance based on the Alveo U30 which is equipped with the new Xilinx Video SDK that makes cloud-scale encoding fast, accessible, and simple to deploy.
Accelerate your streaming workflows with our turnkey software developers kit (SDK). Access the tools, APIs, and high-level FFmpeg frameworks and start streaming in just one day.
Live-streamed video has become essential for work, health, education, and entertainment. Content creators are looking for new ways to engage, and streaming providers are desperately trying to keep up. Faced with costly scale-out deployments that drive up CAPEX and OPEX, streaming providers need a solution that delivers real-time streaming while managing total cost of ownership.
With a Xilinx RT Video Server appliance, streaming providers can now support millions of live user generated content (UGC) streams - no matter how big the audience - with the industry’s lowest cost per channel, highest density, and smallest footprint.
With a flexible, interchangeable architecture, the pre-assembled and configured appliances enable you to deploy on-prem or in the Cloud in as little as one day.
Built for performance & efficiency
Certified High Channel Density streaming appliances available now from Boston and Hypertech
Take the guesswork out of accelerating your streaming applications and leverage the hardware-accelerated features of appliances based on Xilinx technology with our turnkey software developers kit (SDK). Quickly transcode your live video content with standards-based FFmpeg and get up and running in just one day. No FPGA experience necessary.
Or tap into the Xilinx™ Media Accelerator Library, a C-based API that enables you to integrate your own proprietary framework.
Xilinx Alveo U30 accelerator cards handle the entire transcoding pipeline, including ABR scaling, reducing CPU requirements and increasing channel density -- reducing your total cost of ownership by 80%.
Some video transcoding solutions can only handle decode and encode on the card itself and utilize highly inefficient and costly CPU resources for ABR scaling. This forces the deployment of a higher quantity of more expensive servers to handle a given workload, forcing TCO through the roof.
Competing GPUs allow you to do ABR scaling on the card – but you end up burning cores –forcing you to lower your channel density, consume more power, and add more servers to your datacenter.
Xilinx-based appliances support up to 56 channels of 1080p60 and provide an 80% cost savings -- along with 1/10 the power and 1/4 the space.
Industry’s lowest cost & power, smallest footprint
Hypertech-Ciara AP320 server comparison vs. NVIDIA T4 powered server
|Description||Xeon Bronze 3104 @ 1.7GHz||Xeon W-2145 @ 3.7GHz|
|# of cores / Threads||6 / 6||8 / 16|
|Video Transcoding Using CPU Scaling
Throughput for ABR Ladder(s)*
|4 x ABR ladder @ 53fps
(Utilizes 80% of cores)
CPU Cost per ABR Ladder = $223
Utilization per ABR Ladder = 1.2 Cores
|4 x ABR ladder @ 60fps
(utilizes 6% of cores)
|Xilinx – Throughput for ABR Ladder(s)*
||8 x ABR Ladders @ 60fps
(Utilizes 34% of cores)
CPU Cost ABR Ladder = $27.87
Utilization per ABR Ladder = 0.75 Cores
|8 x ABR Ladders @ 60fps
(Utilizes .6% of cores)
Join our subscriber list for the latest updates on our Live Video Streaming solutions.