ATM alive and well

How often have you wanted to go back and check the predictions of analysts and the press for accuracy? Many of them would be hilarious, I’m sure.

Well, here’s one, but it’s not for laughs. It’s a report from Cambridge, Mass.-based Forrester Research Inc. about wide-area Asynchronous Transfer Mode services dated April 3, 1995. The headline: “Wide-area ATM stalled until 1998.”

Take a bow, Forrester. For once, someone got it right on the money.

After years of disappointment, ATM as a wide-area user service is taking its place at the table. Many carriers report that ATM growth far outstrips that of frame relay. Users are more likely than ever to say they’re installing ATM interfaces at data centres and major branch sites.

One of the most prescient things Forrester’s 1995 report pointed out was that for all its multimedia capability, ATM’s growth was dependent on the data network needs of corporations.

Back then, and to some extent today, the critical applications requiring ATM’s even flow of fixed-cell-length datagrams were lacking. At mid-decade, it was becoming clear that videoconferencing, for example, was just not becoming mainstream enough in the corporate world to justify a widespread migration to a new WAN protocol.

Instead, it’s been the simple crush of data applications that have led users to move to ATM because it’s the state of the art today for T-3 and higher corporate links.

In the early days, ATM was supposed to be the follow-up to frame relay. Now it is common to implement ATM with frame relay. For example, if a carrier offers frame-relay-to-ATM interworking, sites requiring 56Kbps to T-1 connectivity can use frame relay-equipped customer premises equipment (CPE), while T-1 to T-3 sites can use ATM-equipped CPE.

Journalists can see this shift reflected in the way ATM is presented by its evangelists. In 1994 and 1995, ATM applications were invariably pictures of the ocean floor, rotating X-rays of sick patients’ innards and other new-day-dawning presentations meant to wow politicians as much as anyone else. Today users almost apologize for how boring their ATM applications can be.

We should be clear what we’re talking about here. ATM hasn’t taken over the world. These aren’t usually people who have brought ATM to the desktop. And the adoption of ATM in corporate networks is taking place on a different path than the raging debate between ATM and pure IP in carrier networks.

In addition, the voice story has changed. Voice has indeed turned out to be a boon to ATM, as corporations have become comfortable taking their internal phone calls off the public telephone network — so long as they don’t replace it with another type of toll system, such as certain voice-over-IP offers.

So welcome back to the fold, ATM. Now if the carriers don’t price it out of the reach of the normal user’s pocketbook…Well, that’s another story.