The Advanced Telecommunications Computing Architecture (ATCA) is seven years old and is now well established as an embedded computing architecture in many different embedded applications. The architecture has grown-up a lot since its early days when analysts were predicting that ATCA would rule the telecoms world and grow to an over US$10bn market.
The architecture was more raw potential than an actual material market back then. Those heady projections from ATCA's early days were never realistic for an embedded computing architecture and so while the more than $10bn market predictions unsurprisingly never came true, ATCA has nonetheless made great progress and established itself as one of the strongest embedded computing architectures. ATCA now represents one of the largest embedded market segments and is projected to reach the billion dollar mark in the next two years – a very impressive feat within the highly fragmented embedded computing market which is comprised of lots of different architectures each serving its own small niche, see Fig. 1.
Core communications
The largest vertical market for ATCA remains by far core communications – the applications at the core of communications networks. Think central office. Core communications applications are exactly what ATCA was invented to serve, so it is not surprising that these applications consume the vast majority of ATCA systems. ATCA is a high bandwidth, high-speed architecture designed specially to meet the demanding requirements of next generation communications applications. It is a perfect fit in central office-type communications infrastructure equipment that sits at the core of all communications networks, and this exactly where most ATCA equipment is deployed today.
ATCA has done particularly well among the tiers two and three network equipment providers (NEPs), who tend to be smaller companies or even startups. These firms do not have the vast development resources and money that their larger tier one competitors do to build their own proprietary compute platforms in-house as part of the end product they are producing. Instead using the standardised, COTS ATCA platform as the compute platform in their products allows the tiers two and three NEPs to outsource this undifferentiated value and focus their own precious development time and money on their differentiating value-add.
Implementing ATCA allows these firms to conserve their valuable development resources for only the most important activities, use the same compute platform across multiple product lines and accelerate their time to market. Most of the tier one NEPs are also adopting ATCA to take advantage of these same value propositions, although they are migrating off of their proprietary platforms much more slowly than the more nimble tier two and three NEPs.
ATCA has grown along with the deployment of converged all-IP networks finding a home in much of the latest communications infrastructure equipment built by tiers one, two and three NEPs. Typically ATCA is deployed in greenfield applications when brand new infrastructure equipment is designed because it is too costly to design into legacy equipment and does make sense commercially. Therefore we are only just entering the period with perhaps the greatest potential for ATCA deployment: the rollout of next generation 4G and LTE networks. The construction of new LTE and other 4G networks will require all brand new equipment, exactly the type of greenfield equipment that implements ATCA. With so much new communications infrastructure equipment being designed and deployed for 4G networks, it will create the greatest potential demand for ATCA since the creation of the ATCA standard in December 2002, see Fig. 2.
Edge communications
More recently ATCA has also found its way into edge communications and even enterprise applications. Smaller and less robust ATCA rack configurations such as two and five-slot systems recently released by ATCA suppliers are a good fit for these less robust applications that do not require the same high performance as core communications applications that require the typical 14 or 16-slot ATCA rack configurations. By taking out much of the performance and cost that is not required for these applications, suppliers have made ATCA a fit in applications for which it was previously overkill.
Core communications will always be the primary and largest market for ATCA, however the amount of ATCA equipment adopted by the edge and enterprise segments is likely to increase over time.
In the army
Perhaps the most surprising vertical market to have adopted the communications-centric ATCA architecture is the military and aerospace vertical. Although when you think about how ATCA fits in mil-aero it shouldn't be. Militaries have to build and maintain their own secure communications networks that in the end are very similar to the same networks built for commercial telecoms. Since ATCA is a high bandwidth architecture built specifically to serve communications applications, it is a perfect fit in military communications infrastructure equipment and this is where ATCA is being adopted in the mil-aero segment.
Much of the mil-aero ATCA deployment is in benign environments outside of the war theatre in communications infrastructure equipment that is housed in say the basement of a building on a military base outside the theatre. However, some mil-aero ATCA deployments have been to the field as well. In addition to deployments on mobile military bases there are examples of ATCA on naval ships and strapped into airplanes.
Similar to the core communications segment, the greatest potential for ATCA adoption in the mil-aero vertical is only just beginning. Many of the military services in the USA are beginning to rebuild their entire communications networks because their legacy infrastructure consists of a mess of stovepiped networks that are different from one another, inconsistent and cannot communicate or interoperate with one another.
Programmes such as Canes in the US Navy and Netcents II in the US Air Force are multi-billion dollar programmes aimed at building new unified, consistent and interoperable communications networks. ATCA could be the potential perfect fit as the computing platform of choice for these programmes. This potential is not limited to only the US market as militaries in Western Europe are undertaking similar programmes to build their own updated and interoperable network infrastructure as well.
Bright future
It is clear that ATCA will not become the computing platform of choice in every single vertical segment across the embedded market, but within the segments that it does serve ATCA is clearly a leading architecture. The future is bright for ATCA and it will likely be deployed in even more communications and mil-aero applications in the future.
Eric Heikkila is director of embedded hardware and systems for VDC Research Group