The Future is... Hardware
In a second blog Andreas Hegers takes a step back from software revolution
Really? What's going on with this guy, I hear you say. In his last blog he’s praising the beauty of standards, and now he’s going completely retro? Hardly anyone is talking about anything but the endless opportunities of software these days. It’s even difficult to find a decent webinar that talks about Hardware … how dare I?
Take a deep breath… and if it makes you feel better, just think about the fact that there’s already a lot of software residing in hardware. But this embedded software has a fundamental advantage over what most people think of as software these days… someone was obviously prepared to pay for it, since it… well, what can I say… sits on hardware. Funny, isn’t it?
Since the advent of open source in telecoms, customers have increasingly fallen in love with the idea of getting all software for free. That keeps vendors scratching their heads when thinking about long term survival, let alone midterm profitability… and ways of differentiating. Hopefully the industry will find a way to keep on paying for software innovation rather than expecting it to happen ‘for free’, based on wishful thinking or long term relationships.
It must be quite a shock for some vendors to realize that service providers are not buying from us for the fun of it. No they are buying solutions only because they a) need them and b) don’t get them for free. Our customers’ main business is serving their customers, not entertaining vendors. Rather, they would live without us and save a lot of money. Brutal, isn’t it? Well, why would they be different than us?
But back to the point… which is the area where you need more than some laptops and an internet connection (yes, exaggerating… :) ) to create value? It’s good, old hardware. What enables networks to transport terabits of information via a pair of fibers? Not open source software, I am afraid. When will white boxes be able to replace something like a core router, let alone an OTN switch or even a teeny-tiny transceiver module? Or build meshed networks with very high reliability?
The key issue of our industry appears to be that we’ve become so afraid of looking old school that some of us don’t even talk about our unique strengths and expertise any more… just because of hipster concepts that won’t work in most telecom networks any time soon.
Really? Hmm, let’s take a look at the MSO space… at Angacom, the largest German cable show, there wasn’t any sign of free software and white boxes. What's so different there? Maybe they are closer to the end-customers, maybe their solutions require more “physics”, maybe they just have less people with time for making up fancy concepts… or maybe they are just far away enough from IT people taking over.
But even in hardware land there is wide open spaces to maneuver and improve: We are approaching the Shannon limit, so we need to find new ways of transporting bandwidth (or just light another pair of fibers…:) ). We could start collapsing functions where it makes sense, rather than building multi-layer networks. We could start building more modular solutions without the need for external cabling and complex maneuvers on NMS level. And the list goes on… lots of work yet to do at ‘home’, with many real, potential customers out there, across all tiers.
These customers are expecting us to leverage our strengths, focus on what makes sense and work with them. Software plays an increasingly important role, so going back to silo concepts based on monolithic models has zero chance of success. But networks will always be built based on a mix of software that works “standalone” plus software which is embedded in “boxes”. The absence of the latter would make virtually all service provider infrastructure disappear in a heartbeat, except for very simple topologies.
So let’s continue working on open, futuristic concepts, but let’s keep in mind that most customers want solutions that will improve their business straightaway, rather than those that require a complete overhaul of the network and processes for them to work. And from this perspective, both software and hardware are still on a par when it comes to value going forward.
The hype of newer software oriented evolutions is far from being lived up to. For example have a look at this webinar by industry thought leaders. A joint ECI and Light Reading Webinar with guest speakers from Level 3, Orange Business, and LinkedIn: Has SDN Lived Up to the Hype.