Does the cloud have a silver lining for TV?

The cloud has seemed to be an almost magical phenomenon in the software industry during the past 10 years or so. Migrating to the cloud was the hot topic back then, and is still one today! It seems that everything we used to have on-premise has been moved to the cloud.

Recently, Netflix announced that it has now transformed into a “cloud native” company – after seven years migration from data centers to the cloud. The cloud migration has reportedly resulted in 1000x growth in streaming hours over the years!

So, surely the cloud is here to stay. The elasticity of the cloud and virtualized environments enable streaming services that can dynamically expand and shrink with demand, eliminating wasteful over-provisioning. Cloud is particularly useful in these scenarios:

  • Initial deployment – your engineering or devops team can instantly spin up instances to pilot or try out new services in minutes
  • New market expansion – you can test the water by expanding (or shrinking) your current footprint based on demand
  • Serving volatile markets – you do not have to worry about adding or removing capacity in a market where demand fluctuates (resources are automatically adjusted)

And surely the same is true for delivering TV? It certainly makes it easier for you if you don’t own any network or POPs to deploy your own origin or CDN solutions and for those who wish to replace capital investments with operational expenses. Moreover, this model doesn’t normally require a huge change or investments in the technical organization and competencies.

However, if you are delivering TV services – and if you care about delivering an amazing experience – simply locating everything in the cloud is not going to cut it.

Cloud serves some functions perfectly. But it simply doesn’t make sense – from cost and scale perspectives – to move traffic intensive functions such as storage and caching into a central location. They can be virtualized. They can be run in software – or they can be run on hardware accelerated servers. But they need to be located closer to the viewer to deliver cost and scale benefits.

Cloud and virtualized components are best complemented by distributed TV delivery based on TV optimized network appliances – especially in the following scenarios:

  • High capacity deployment on a given footprint – with deterministic performance, you can sustain video transmission performance at wire speed bringing a significant improvement to viewing experience
  • Highly distributed network deployments – the server is built to last – with no moving parts – which means you can place the server far out at the edge of your distribution network
  • Many TV services – you can enable a variety of TV services on a single appliance without installation of additional hardware or changing the system picture

For those with an existing network, distributing TV servers at the network edge give the best cost advantage by the ability to deliver both OTT and IPTV on the same platform. Most importantly, viewing experience is highly affected by how close your content is to your viewers. Edgeware has learnt this lesson through hundreds of Tier 1 deployment of TV services with major players in the broadcasting, telco, content and cable industries.

In summary, cloud is here to stay, however it will not be alone. And it’s not the magic solution for TV delivery.

Even “cloud native” Netflix has distributed purpose-built servers to create its Open Connect service in ISP’s data centers to improve performance and reduce the capacity ISPs need to build.

A wise colleague of mine once told me, “Behind every cloud, there’s a great deal of infrastructure.” The silver lining for TV delivery, is that this infrastructure does not need to rely on magic.

PS: You can always follow us at #thereisnomagic on Twitter. If you enjoy this post or contemplating on TV delivery, don’t forget to tag us.



Ling Koay
Product Marketing Manager


Fill out the form below and we will get in touch with you.