As the technology world grapples with the impact of Steve Jobs’ resignation, many journalists and analysts are in a reflective mood, pondering Jobs’ legacy and achievements at Apple. One small part of this legacy will be Jobs’ role in driving the continued video explosion. Jobs acted as a key enabler in creating almost ubiquitous access to video, both in regards to consumption and to sharing. One need only look at the amount of YouTube content viewed on iOS devices to understand the figures involved.
What’s incredible to consider is there are no signs of video demand slowing. Cisco’s Virtual Networking Index (VNI) highlights this continued spectacular growth. Indeed, by 2015, Cisco expects that video traffic will account for 90% of all consumer Internet traffic. What’s more, much of this video will be High Definition (HD). According to the research, 77% of Video-on-Demand (VoD) will be broadcast in HD. Also, VoD traffic will itself increase by 300%, equivalent to 3 billion DVDs downloaded per month.
Yet what I find fascinating about these predictions is that they predate any real understanding of traffic on tablet devices. In an earlier post, I suggested that we’re standing on the precipice of tablet data consumption. The traffic we’re seeing today will dramatically increase as tablets become more affordable and widespread. As this happens, video on tablet devices will continue to rise, especially as applications such as the BBC iPlayer and other such streaming services mature.
The expected growth in video traffic is something that most service providers are closely monitoring, but GigaOm recently uncovered a joint academic paper from researchers at Akami, Harvard and the University of Massachusetts that paints an alarming picture of what the explosion in video traffic means for the network. Akami notes that President Obama’s 2009 inauguration drew 7 million streams demanding almost 2Tbit/s of data. However, this is only a fraction of what’s to come. The paper highlights that by 2015 video events will generate upto 100Tbit/s of data. This is the equivalent of delivering a TV quality stream to a large primetime audience.
The paper continues by addressing what can be done to meet this demand, suggesting that the bottleneck is no longer the origin data centre and that capacity needs to be developed throughout the network,
“One must consider the throughput of the entire path from encoders to servers to end users.”
Without question, this is one of the most critical elements to meeting tomorrow’s video demand. Capacity needs to be available throughout the network.
Many service providers are aware of this situation and are at the first stages of rebuilding their networks to accommodate the forecasted growth. Much of this rebuilding is part of the optical reboot and primarily focuses on the rebuilding of our global core networks on a foundation of 100G, OTN and ROADM technologies. Since my previous blog post on the topic, we’re starting to see deployments gather pace and it’s incredible to consider the capacity that will be available.
However, it’s critical that we don’t only develop the core. Access networks are also a vital element to meeting this bandwidth explosion, whether it is fixed line or mobile and there’s a great deal of work to do here. I’d be interested to hear your thoughts on video's continued explosion and if you believe that our networks are developing at the speed required to meet the enormous bandwidth challenge ahead.