The implosion of token ring

With the near simultaneous implosions of IBM's Networking Hardware Division (NHD) and Olicom at the end of August, any remaining hope that token ring would continue as a strategic network technology was snuffed out.

Victim of a deadly downward spiral of lessening customer interest followed by correspondingly drop in vendor interest, token ring has "augered in." Olicom, whose stock was trading at 11/16 the last time I checked, cited dramatic decline in token-ring sales as the cause for the company's accelerated demise.

Ironically, there are still many customers out there who've got networks that can continue to run fine for years to come. But, like it or not, they will have to begin the forced march to Ethernet.

Just a year ago, things looked very different. The religious wars were over and Ethernet was victorious, but token ring was clearly a viable niche market. While greenfield sites would likely be Ethernet, it was still usually cheaper to buy the more expensive token-ring network interface cards (NIC) and switch ports to update and extend existing token-ring LANs than to begin a massive migration to Ethernet.

When it came to advancing token ring to the 100M bit/sec level, the field had narrowed to three: IBM, Madge and Olicom. But to many existing token-ring customers, the IBM/Madge/ Olicom troika represented more than enough commitment. Many customers were anxiously awaiting High-Speed Token Ring (HSTR) NICs and switch ports that were slated to ship around year-end 1998.

More than that, though, before committing to 100M bit/sec token ring, customers were waiting for IBM et al to announce an aggressive campaign to make up for lost time and produce industry-standard Gigabit token ring in 1999. Clearly, they wanted that message as a guarantee that money invested in HSTR would not simply be good money after bad. What was the answer to the Gigabit question?

As my German friends would say, "Keine Antwort is auch eine Antwort." Rough translation: No answer is also an answer. IBM's failure to take a positive, aggressive stance no doubt sent a clear, negative message to many of its larger customers. If IBM isn't bullish on token ring, they likely told themselves, it is time to put our Ethernet conversion into high gear.

Last December, IBM quietly cancelled its announced HSTR switch port module - leaving only Olicom and Madge to deliver on the promise of HSTR switching. While both companies had churned out impressive technology over the years, by themselves, they could not sustain an industry.

Today, Olicom is gone - having sold all its token-ring technology to Madge, and IBM NHD is just a bit player. That leaves Madge the undisputed "premier" purveyor of token ring in the world. And what is Madge doing? Trying to become an application service provider as fast as it can. Convert? No longer is it a question of "if," but merely of "when."

More Tolly on Technology columns