Room At The Bottom:The Promise Of Micro-TCA
by Lee H. Goldberg
One of the basic laws of networking ecology states that potential markets and sales volumes for equipment grow as they get closer to the end user. So while the ATCA standard will enable great scales of economy in the next generations of major enterprise and carrier-class equipment, it will take another standard with more modest goals and form factors to work the same magic at the network edge. And if I'm reading the smoke signals I'm seeing on the horizon correctly, the emerging micro-TCA (µTCA) standard will emerge as the platform of choice for everything from DSLAMs and media gateways to right-sized IP-PBXs.
In fact, the term "right-sized" might be the most appropriate way to describe this emerging standard that uses its parent's smaller advanced mezzanine cards (AMCs) as the basis for a family of smaller, less expensive chassis configurations that still pack up to 750 Gbit/s worth of capacity in their backplanes. The µTCA's mission appropriate form factor will also still provide all of ATCA's critical shelf and system management features, as well as the option for redundant power.
Initiated in 2004, the µTCA standards effort is enjoying industry broad support thanks to its promise to enable lower cost equipment that leverages the technologies and manufacturing base already developed around the rapidly-maturing ATCA spec. I saw the first public demo of a µTCA chassis this spring at the 2005 Supercomm in Chicago, and while the lash-up on display lacked some of the refinements a production system would have, it clearly illustrated what one could expect after the µTCA spec is finalized some time in mid-2006
µTCA achieves its smaller form factor by eliminating ATCA's shelf management connectors, shelf management controller and hub cards, and uses much smaller carrier cards to hold the AMCs. The current version of the proposed standard calls for a virtual carrier concept that holds up to 12 AdvancedMC modules, and a half-height variant for use in even more compact products. Each carrier can be mixed-and-matched within four form factors, including a compact cube for small, non-rack applications.
This and a simpler, smaller power supply scheme lowers µTCA's overall cost and enables a much lower price entry point while still delivering carrier-like reliability. In addition, its 300-mm depth form factor (vs 450-500 mm for the full ATCA chassis) makes it a better fit in the tight confines of a DSLAM or other pedestal-based access box. It's also more compatible with the shallower telco equipment racks being used in Europe.
With a SerDes-based backplane that's scalable between 12 and 750 Gbit/s, µTCA allows ATCA technology to reach many markets at the bottom of the food chain which were simply not cost effective before. While the smaller chunks of the compute power and bandwidth available in 1st-generation products (expected in early 2007) will be a stretch for even smaller RNCs, they'll be ideal to meet the requirements of DSLAMs, IP-PBXs, media and signaling gateways, as well as less obvious applications like SANs and other storage systems. I even expect that the economies of scale generated by the high volumes in these markets should "trickle up" to lower the cost of equipment based on the full-sized ATCA chassis.
Much like the standard PC platform has become the basis for a robust and diverse industry that provides the majority of computing elements we use, the ecosystem created by ATCA and its new junior partner µTCA will benefit the telecom/datacom industry with an infusion of lower-cost, open-architecture equipment. While doing business in the highly-competitive open marketplace that standards-based platforms create may be a challenge for some members of the old-school big iron gang, the innovators who will provide the real direction for the industry will be sure to thrive.
Comments? Questions? Hot tips on important industry trends? Write me at: lgoldberg@green-electronics.com