High-Density Transcoding with Wowza and AWS VT1 Instances

The more you scale a system, the more designing for efficiency matters. Wowza’s recent launch of Wowza Streaming Engine with the AWS VT1 instance is significantly more efficient and cost effective than regular cloud based live video solutions.

As impressive as that sounds, there’s more to the story. Running Wowza Streaming Engine on AWS VT1 isn’t for everyone, but if you’re running live video transcoding at scale, this could bring you savings, operational simplicity, and more.

 

Who Benefits?

WSE supports a broad range of compute options on AWS. Wowza has tens of thousands of Wowza Streaming Engine instances running across AWS environments, supporting many different use cases and customers.

Our cohort of users come with specific demands. Many are looking to process high volumes of transcoded streams – typically a minimum of around 50. This is often the case with:

• Online Video Platform providers looking for multi-tenanted cost optimization
• Content Owner operated video platforms fine tuning their infrastructure
• Surveillance systems with dozens of camera feeds requiring transcode
• Machine watched video processing with little streaming
• User Generated Content from many sources

 

What Makes VT1 Special?

Why are we making a fuss about this AWS VT1 instance? Put simply, it’s the right tool for the right job. We have GPUs and CPUs, but VT1 is based on AMD’s U30 card – hardware specifically built for high density transcoding. Wowza has worked closely with AMD to optimize this integration and make the most of these unique capabilities.

VT1’s hardware includes dedicated GPU chips for transcoding alongside CPU processing. Therefore, whilst we talk about ‘transcoding’, there’s more to it, especially with Wowza Streaming Engine. With WSE’s packaging and media server (origin server) capabilities to hand, the CPUs are fully available to perform. This combined solution also effortlessly supports CRM and AI processing integrations.

Essentially, the VT1 hardware architecture with Wowza’s dedicated integration solves two problems:

1. Inefficient use of GPU resources per card
2. Maxed out CPU’s

Instead, we see optimal distribution of transcoding and video processing across the VT1 instance. This means cost and power savings, together with simplified architectures. At scale, this is significant.

 

VT1 and Transparent Pricing

Cost effective streaming is about more than just systems level design efficiency. Wowza’s VT1 pricing is itself designed to help streamers save money by offering a predictable total cost of ownership (TOC).  

Typical cloud based live encoding pricing options often charge for transcoding alone, separating packaging and origin functions. They also frequently charge per input stream and per output stream separately. Long story short, it’s difficult to predict how much you’ll be spending overall and easy for you to suddenly be spending more than you planned.

Wowza’s pricing is simple, easy to understand, and easy to predict as your streaming needs scale. We charge per hour, per instance. That’s it. At 100 channels, WSE on VT1 is ten times, or 90% lower cost than transcoding only on other cloud-based solutions.

 

Conclusion: Is Wowza and VT1 For You?

Wowza and VT1 offer substantially lower cost at scale, including packaging and delivery. They also offer simpler architecture while still being feature rich and scalable. In short, it’s an ideal solution for organizations that need to scale on a budget while still maintaining some control from the on-premises build through the audience experience.

Learn more about Wowza Streaming Engine software and AWS VT1 instances by booking a consultation.