I commented on a blog post earlier and I am recording my updated thoughts to review over the next few years to see how wrong I was. My initial business cases still have not changed much from 4 months ago.
It has always been and will be about money. What will drive this is the ability to commoditize networking vendors and level the playing field, creating competition and bringing some parity between vendors and price points.
Everyone who is in the networking industry today that is reluctant to admit their hardware is a commodity on some level, will fall back to their software being the value. Smaller vendors will exploit the opening and take market share from the big boys by pushing eliminating the software value with standardized firmware. Self-provisioning as a services will drive this also. Why build regional networks when i can buy a pre-built statewide/nationwide network slice and provision my own policy as the business requires.
In the data center the value prop is the ability to abstract network into software right along with storage, hypervisor server etc. Let’s have the same guy that can point and click a Vmotion migration from one data center do it all rather than adding another high paid operator to nail up a VPLS/OTV pw/vcircuit and avoid the overhead all while keeping it all agnostic to the hypervisor, vswitch, network underneath.
Speeds and feeds will get there if the vendors decide they can make money on a decoupled control plane.
What will be the revolutionary technology becuase we are pretty short on them over the past decade. OSPFv13, L2TPv9, TRILL (assuming it is ever ratified and supported on your lifecycled frankensteined big expensive hierarchical replaced line card data center switches)? Its kind of embarrassing 10-12 years later we still run STP (granted non-blocking) in our data center.
Maybe I need more hype but I am willing to bet on seeing more innovation and change over the next 5 years rather than another 15 years of the same.