John Gallant
Network World editorial director
ATM: Requiem for a dream

ATM was undoubtedly a flop - a local-area networking dead end for corporations that committed millions of dollars to implement it and for many of the start-ups that were launched at the height of ATM mania in the mid-'90s. (Although some of those newcomers did quite nicely for investors, thank you, including Fore Systems, which Marconi acquired for $4.5 billion just before a much bigger bubble burst, the dot-com one.)

ATM wasn't some run-of-the-mill failure; it represented the death of a grand dream of unified networking across the LAN and WAN. ATM was envisioned as a single network technology that gracefully would support converged applications across enterprise and service-provider networks.

ATM has had an honorable career in the latter, but it never achieved the hoped-for glory in the LAN, where it was tripped up by its own complexity and technical limitations, and the relentless improvements in Ethernet's price and performance. Like the Open Systems Interconnection model, ATM takes a special place among other honorable efforts to make the network world a better place. For that, a moment of silence, please.

Did we get it right? Discuss our choices in our flops forum.