Intel bought Aepona and many onlookers have been wondering why. It strikes me that it might make a bit more sense in the context of Oracle’s recent acquisitions of Acme Packet and Tekelec.
It’s important to keep in mind that for a behemoth like Intel, there’s not much risk here. Aepona’s $40MM in revenue is a drop in the bucket for a company that does more than $12.5 billion per quarter. The word on the street is that Aepona will continue to operate as an independent entity; if that’s true, then Intel doesn’t even need to go through post-merger pain. Intel can afford to incubate Aepona and see whether the company is a visionary that is ahead of its market or will end up being a small write-off on a quarterly report.
Aepona is one of several players that has tried to bring the concept of Network-as-a-service (NaaS) to market in a telecom environment where the idea of opening network resources to developers goes against all convention. I’m just old enough to remember when AT&T (the original monopoly, not today’s re-branded SBC entity) manufactured and owned every telephone attached to its network because it felt that third-party equipment would introduce too much risk. So, if that’s basically our starting point, you can understand why telcos have not largely jumped onto the NaaS bandwagon yet. Aepona has been pounding the pavement trying to get anyone who will listen to understand that the world has changed. In an environment that is all about mash-up applications and fostering third-party innovation it’s hard to find a deeper resource platform to leverage than telco networks. The concept makes sense, but we’ve often discussed here in BillingViews that the language gap between Hackers and Engineers is often a problem, on top of telco inertia and wariness of uncertainty and radical change.
So, what does this have to do with Oracle? Well, NaaS kind of falls into this new, intelligent layer we see emerging above the network. Oracle bought session border controller vendor Acme Packet and then diameter signalling control provider Tekelec right on its heels in part to carve out ownership of what could be a new market segment (and will be if and when any major research firm decides to track and promote it as such…Shira Levine is doing really good work here for Infonetics). So, Intel – from a telecom industry vertical point of view – is jumping into it as well, in a way that makes sense for Intel. Intel needs to add-value on top of processing, which is already commoditized, and to dominate as much of the cloud-computing world as it can. But, if you’re going to dominate cloud-computing, you can’t just sell more commoditized processing cycles. You have to sell stuff that generates an awful lot of transactions and functionality that in turn drives cloud-computing resource utilization. NaaS arguably fits that bill nicely, if Intel’s weight behind it makes it take off.
If we move to an environment where NaaS is widely adopted, the model could dovetail right into all of the insane stats we’re seeing around the growth of digital content and services; M2M or IoT; more complex SaaS models; and anything and everything that involves real-time transactions, from autonomous vehicles to surgery performed remotely via robot. NaaS matters to all of this because it, in theory, opens up access to connectivity, security, QoS and QoE, policy-mgmt, routing, and all of the network-intelligence kinds of things like, frankly, companies like Acme Packet and Tekelec enable. This is a vision of an industrial-grade, large scale mash-up environment that is talked about in an abstract sense, but isn’t yet a reality.
So, it may be that Intel sees this vision and figures buying Aepona is a small bet it can place with a potentially huge payoff for its future business. And, if the thing that has held Aepona back from driving NaaS into the market was that it wasn’t big and rich enough on its own to a) promote the heck out of the idea, and b) overcome Telcos’ extremely conservative (and legitimate) concerns about risk, well, Intel’s deep pockets, market power, and visibility could change that.
What I think is a very different challenge Aepona will face than Acme Packet or Tekelec, however, is the Hackers v Engineers language gap. It will take some highly intelligent and creative story telling to get the world – and developers from the Hacker world – to understand why NaaS is valuable to them, while at the same time convincing Telcos that it’s a money maker . So, this is a sort of two-sided business model where neither side has quite matured or become accustomed to working together yet. We’ve seen companies like Voxeo Labs and Twilio have some success in closing aspects of this gap, but now it’s up to Aepona – with Intel’s cash behind it – to learn from that success and then replicate it on a much grander scale.