Ben Weekes is a Senior Architect at Agora. An original pioneer and innovator of WebRTC technologies, Ben was Founder and CTO of Requestec which was acquired by Blackboard in 2014 for expertise in WebRTC conferencing. After serving as Chief Architect at Blackboard, he joined Agora to continue his journey in pushing the boundaries of video telephony.
I recently debunked several of my own misconceptions about real-time-communication during my Kranky Geek WebRTC presentation. I’ve been working in RTC for nearly a decade, but only in the last year have I come to understand that the following commonly held beliefs are simply not true for video chat and streaming:
- Peer-to-peer connections between two people provide the best results
- The public internet is faster than a virtual network
- Single selective forwarding unit (SFU) architecture (server/router relays) is best for small groups
In my talk, I demonstrated that multiple SFUs, connected by a software-defined and managed virtual network, can significantly outperform peer-to-peer connections over the public internet. This performance advantage means a more predictable, noticeably higher-quality, real-time video experience for participants.
I don’t want to give away the entire presentation, but the gist of it is this: On the public internet, you have no control over routing. Each participant could be using a different ISP and each of these ISPs is routing traffic based on cost—not efficiency—in this order:
- Their own network
- Peer ISPs with free pairing agreements
- Least expensive higher-tier option
Because of this business-relationships-driven model, it is difficult, if not impossible, to ensure that each participant is seeing and hearing the same thing at the same time during interactive video chats. If you have ever been on a video call where there is noticeable delay, you know how counterproductive this can be. Agora has engineered a unique solution to this unfortunate reality. We have placed our own servers (SFUs) into every ISP network around the globe and we use machine learning to continuously probe for the most direct and robust routes between them. Instead of relying on the public internet, we proactively manage the communication from end to end to ensure the highest-quality experience for everyone—regardless of location.
On average, the Agora architecture provides twice the speed of the public internet—allowing all participants to see and hear things at the same time—in very high fidelity. But this is not the only benefit: This unique architecture is infinitely scalable! We have tested with up to 1 million people in a single video conference with everyone experiencing ultra-low latency (delay) and high-fidelity.
Agora’s reliable, very high-quality, real-time engagement (RTE) platform provides worldwide, scalable, interactive video streaming delivery. And since we’re a developer-first API solution, we make it easy for developers to embed voice and video chat or streaming into any app using our easy-to-use native and web SDKs.
Check out my Kranky Geek presentation—including a real-world test—below.