Just catching up with this announcement from last week: Cisco briefed analysts and press on a product that they'll be showing at VoiceCon, and which is, in fact, their entry in the Finals of the Best of VoiceCon awards: The Inter-Company Media Engine, or IME.When Tony Bates, who's also going to be a VoiceCon keynoter next week, led the IME briefing, he didn't shy away from calling the product a "disruptive, important piece of technology," despite the fact that, just two days earlier, Cisco came in for a lot of criticism for the extravagant hype that they promulgated for their "change the Internet forever" announcement of the massive CRS-3 router.
Ultimately, when it came to the router, I think the marketplace reached the same conclusion that Zeus reached in the post I linked in the previous paragraph: The CRS-3 is truly a major announcement, but Cisco just oversold it, using Apple language to describe a product that only the most technical followers could possibly view the way the broader market looks upon Apple products.
In the case of the new IME, the claims were a bit more modest but still aggressive, and the product really is a significant piece of technology. I think the easiest way to understand it is to look at the diagram below, which is a little complicated when you can't do it as a build, but is still reasonably clear:
What happens is, the first time the person at Company A calls the person at Company B, the call goes over the PSTN--i.e., the route up at the top. When that call is completed, the call controller (Session Management Edition) sends call detail info to the IME boxes that sit either within the enterprise or a service provider network serving an enterprise. The IMEs on either end of the call then exchange information and register a SIP route across the IP Network so that future calls can be routed there rather than over the PSTN.
What all of this means is that you can call someone at another company and establish SIP-based communications with that person, without the end user (or even the network administrators) going through any pre-configurations. And it doesn't just apply to telephone calls--any media that can be carried over SIP, that an end station is configured for--say, video--can now be supported over that pre-configured route.
So Cisco's come out with another one of their great technologies for doing amazing things--as long as you do it with Cisco gear--right?
Well, not really. They've submitted the technology behind IME to the IETF for standardization, as "ViPR" (Verification involving PSTN Reachability), so ideally, the thing would evolve like Tag Switching did during the 1990s, where a Cisco-proprietary idea eventually became the widely accepted standard known as Multi-Protocol Label Switching (MPLS).
Even with the IETF's imprimatur, I don't know that Cisco's competitors will rush to embrace ViPR. Cross-enterprise interoperability is critical to creating the "network effect" that would drive success for personal or desktop video, which is clearly what Cisco wants IME to do--as John Chambers repeated again with the CRS-3 router announcement, anything that loads IP networks, Cisco is in favor of. If other video-centric vendors like Polycom, decide they'd rather cooperate than fight, they could endorse the idea of ViPR; or they could discourage it. I'm not sure what's in it for an Avaya or Siemens to go out of their way to drive video at the desktop. If their customers start demanding it, they may have to revisit it, but at the outset, with so many other issues demanding their attention, I don't see why they'd make it a priority.
Having said that, are enterprises likely to demand IME-type functionality? Joe Burton and his Cisco colleagues offered a couple of drivers for enterprise adoption: Cost savings by avoiding the public network; and the ability to collaborate with partner/customer/supplier organizations more seamlessly. Of course, cost savings are always interesting, but I don't think the business case for doing this is going to be particularly straightforward.
Architecturally, IME is a pretty interesting development. But I think that a lot will have to evolve internally among enterprises before a lot of them feel like they need it.
