They have some patents and expertise in thin server design that were good enough to make IBM pony up $5.6M in licensing fees a few years ago. IBM invented computing and generates more patents than anybody else so, while largely unnoticed, that 1999 deal to license thin server reference designs from an unknown like NENG, which then had less than $2M in sales, was fairly significant. Perhaps that combination of patents and expertise is one of the reasons that they won the initial Centera contract at EMC, which probably has one of the toughest qualification cycles in technology.
Other companies in the server appliance business are RTX (private) and SteelCloud (public). IBM has farmed out its server appliance business to a top 5 contract manufacturer. Of course, the server vendors like Dell, Sun and HPQ all have the capability to make their own appliances. These vendors, however, focus on server appliances while NENG is carving out a niche in storage and security appliances, which are expected to grow much faster.
If you look at Centera, for instance, the growth potential is simply outstanding. One study already shows that fixed content will represent more than 50% of all stored content by 2005. One Centera customer (hospital) expects to increase its Centera storage 10x in the next two years. Another Centera customer (hospital) just bought 128 terabytes of Centera to house and network all its medical images. Entry-level Centera v1.0 was 10 terabytes (raw), but the average Centera version 1.0 deployment was 27 terabytes (raw). Those early adopter numbers from 150 customers (health, finance, manufacturing) tell you that a customer quickly adds more nodes once a customer learns how to deploy and use Centera. Customers typically buy their storage monthly, quarterly or annually to take advantage of the typical 35%-40% annual drop in storage capacity prices.
Centera v1.0 was rolled out in April 2002 and Centera v2.0 was only rolled out in April 2003 so it's still too early to gauge how long the recurring revenue potential is for EMC or NENG, but I already think this is the EMC product line that will have the longest recurring revenue potential because all a customer has to do is keep on adding Centera nodes to the IP network.
