Fun Fact: A 60-Second Video Is Equivalent to 1.8 Million Words

Well according to James McQuivey of Forrester research in his paper “How video will take over the world” back in 2008, video was set to become the pervasive medium. His prediction has come true, in that today video is the fastest, most convenient way to come up to speed on practically any subject.

Communication has moved on from the age where it was enough to have pages of information on the internet, to a point where 5 billion videos are being watched on YouTube every day. According to the same article – 1.8 million words – is equivalent to 3,600 average informational web pages.

Did you know that the most viewed video of all-time on YouTube is Despacito by Luis Fonsi with 5.8 billion views?

What’s true today and foreshadowed in the article, is that video is one of major workloads that you have to be planning for in a server infrastructure. The rise of Machine Learning and AI along with advances in processing algorithms and power mean that understanding video content and its context is now possible. Video represents a way to learn and understand all sorts of things to do with work, not just music videos on YouTube. Sometimes seeing is understanding a problem. Being able see why a widget does fit - because of the clearance needs to get that widget into place - can be the difference between great customer experience or a poor one.

I’d be interested in hearing your comments – let me know and join the debate below..

IBM Power and New Workloads

As we evolve our systems of record and systems of engagement into systems of intelligence – we need to be able to process video with intelligence using machine learning and AI – this means have the right workload architecture. As this recent Forbes article argues…

Many of the libraries are being developed specifically for emerging applications like image and video recognition. In using the ImageNet dataset, IBM claims that can train neural networks up to 4 times faster than a Xeon-based system. With Caffe, IBM was able to take the training of a neural network from hours down to minutes while making it easier for data scientists to use more intuitive tools rather than learning to code.

Some 10 years on from Forrester’s prediction that video would take over the world – we are now experiencing a significant shift in the need for server processing power to cope with demand for video capture and storage in workplace applications.

Is your infrastructure ready to cope with video workloads?

I’d be interested in hearing your comments – let me know and join the debate below..

Likewise, I’m always up for a chat if you have any questions on how you can optimise your investment in SAP by reducing infrastructure costs or reach me via linkedin

Share this:

Comments are closed.