Roughly 15 years ago, the network was highly relevant.
At the end of the 1990s, networking sat at the head of the table for providing technology that would connect us all with each other, support online commerce, bridge the haves and have-nots with both connectivity and content, etc.
Back then, you heard a lot of very public pronouncements: “Internet traffic is doubling every three months.” Or, regarding the future scaling needs of the Internet, “If you aren’t scared, you don’t understand.”
But even at the peak of the dot-com boom, many regarded network technology as just “dumb pipes” that simply served the function of moving bits from one destination to another. The hope of many in the world of technology was that the network “would not get in my way.” At its simplest, the network was viewed as the “ugly” — or at least less-intelligent — stepsister of key parallel functions of computing and storage.
When the downturn of the early 2000s hit with its unforgiving force, even that fairly simplistic (and really unfair) perception of the value of the network was undermined.
What happened? Absent a lengthy business-history lesson, there was clearly overinvestment in capacity that was never put to work when the promise of Internet 2.0 was pushed out. Carrier business models that were to validate network investment never materialized, falling victim to consumers’ expectations of free services, and challenged by what some would call a harsh regulatory climate.
That promise of the networks — new services and significant ROI — was never realized. As a result, the network and carrier-network investment took a “time-out” of sorts, and perception of its role and importance regressed.
But with the Consumer Electronics Show under way, there is no better time to explore the underpinnings of almost all the “connected” advances on display this week in Las Vegas.
Today, networks are finding themselves at the big-boy table, sitting next to — and ultimately soon to be arm in arm with — the computing and storage functions.
Changes that have pushed the network forward include the advent of software-defined networking (SDN) and increasing acceptance of the cloud; carrier business models that have become more creative and matured (in a good sense); the advancement of Ethernet as a reliable and ubiquitous technology; and perhaps as important as any, the emergence of an app-crazy culture that has changed the nature of consumer demand and expectation around the online (network) experience. (Thanks, Apple.)
If we focus just for a moment on the rise of apps and the network’s role as an enabler and protector of the application’s value, we can see how far the network has come, and the potential still ahead.
The emergence of the network as a performance-on-demand platform for applications, rather than just “pipes,” has spawned and fueled companies like Netflix, Hulu and Dropbox. In these ecosystems, the network is taking on a more specific and valuable role, connecting virtualized computing and storage resources to deliver unique services. This “network as a platform” model is about connecting users to content, and connecting content to content in an application-centric world.
The network’s role here is to manage the quantity and quality of the connection to ensure the integrity of the app toward delivering on the optimal end-user experience. This makes the network more important than ever before — as a more meaningful differentiator to the carrier or enterprise operator, and as a service of sorts that all of a sudden comes into the consumer’s mindset.
With app integrity as just one example, we’re now long past that original network charge of connecting locations (residences and businesses). As mobility has taken center stage, we’re looking in the rearview mirror at that next iteration of connecting people.
Looking further forward is what some call the “Internet of Everything” and others the “Internet of Things.” In short, this refers to connectivity between machines and devices — it’s very real at CES, in the form of devices that report our workouts to our smartphones and refrigerators that alert us when we are running low on milk. Researchers suggest that more than 30 billion devices will be connected to the Internet of Things by 2020.
The demands of this highly-connected world will require even greater intelligence than ever before from the network as traffic will be more dynamic and unpredictable. Some might characterize this network as the on-demand network available to deliver against whatever mission, whatever application, is most needed in any given moment.
The network we are driving toward is just that — one where liquid bandwidth can be had in the same way that compute and store can be purchased on demand. The envisioned network is low cost and offers great scale; it’s fully programmable, offers rapid service turn-up, and is software-defined and -controlled; and is less about traditional geographic domains (like edge, metro and core) and more about functional domains (such as content servers and virtualized network resources). This network seamlessly integrates connect, compute and store capabilities.
This convergence of the telecom (connect), information technology (compute) and data communications (store) worlds marks the ultimate frontier. The last step to realizing the above vision is the network, of course — followed by those fun discussions on standards, openness and APIs between folks from the connect, compute and store camps.
But because the network was last to the dance, it remains the last function that needs to be checked on our way to building a truly better machine that delivers a performance/demand environment.
The good news is that the network matters again. But, perhaps not surprisingly, it’s for decidedly different reasons than why we valued the network just 15 years ago. Sitting to the right and left of compute and store, carriers and enterprise now see their networks as a strategic asset to delivering a superior customer experience and winning in the marketplace.
Stephen Alexander is SVP and chief technology officer for network specialist Ciena. Before joining Ciena, he worked at MIT Lincoln Laboratory, where he last held the position of assistant leader of the Optical Communications Technology Group. A recipient of the IEEE Communications Society Industrial Innovation Award in 2012, he is currently an associate editor for the IEEE/OSA Journal of Optical Communications and Networking. Reach him @Ciena.