Amidst a challenging economy, it has become clear to all businesses that the cost of service or network downtime is greater than ever. Millions of dollars per hour is the cost of outages in key industries. And according to recent Gartner analysis, the outage costs span a wide range of industries beyond telecommunications, including financial services, manufacturing, retail, and energy. This wide range calls for solutions suited to a mix of environments.

AdvancedTCA is a COTS open-standard bladed computing architecture. It can address environments ranging from semi-industrial to data-center-like. From its inception following the burst of the telecom bubble in 2001, AdvancedTCA has been serving as a fundamental basis for network-ready, carrier-grade common platforms.

Telecom success

The AdvancedTCA standard has established itself as the premier technology platform for telecom applications. Initial deployments centered on control and signaling applications, including IMS core, CSCF, HLR, RNC, and BSC. (See sidebar for acronyms.) Next-generation applications have since been enabled, thanks to the advent of 10 Gbps fabric switching technology. 40 Gbps-ready infrastructure is already available.

All of the network equipment manufacturers who deployed in Phase 1 now have second- and in some cases third- generation platforms under development. These deployments are driving application and volume expansion. Comparing AdvancedTCA to other technologies has industry analysts projecting that it will represent the majority of merchant (open market) embedded computing spend for telecom applications over the next five years and beyond.

For proof, we only need to look to AdvancedTCA-based carrier grade products that include:

n Nortel’s Versatile Service Engine – deployed by AT&T

n Alcatel-Lucent’s HSS, SGSN, and IP-BSC – deployed widely

n Motorola Networks’ Carrier Access Point Controller – deployed widely

n NEC’s Radio Network Controller – deployed widely

n Tekelek’s TekCore IMS Core Platform

n Huawei’s IMS Core Platform

Recent analysis (summarized in VDC’s Global xTCA Opportunity report published in March 2009) indicates that the market for AdvancedTCA merchant platforms and blades is projected to be worth around $530M in 2009 and as much as $1.3B in 2012 even despite the effect of the current recession. Furthermore, the ecosystem now extends to more than 50 main suppliers and more than 100 related suppliers, providing Network Equipment Providers with a combined portfolio of 360-plus AdvancedTCA products. As a result of this unprecedented growth, as well as the intrinsic flexibility of the standard, AdvancedTCA technology is now being considered for multiple applications outside the Central Office, where a more relaxed cooling environment exists, as lower operating temperatures facilitate higher performance potential.

Server-class opportunities for AdvancedTCA grow

Thanks to this well-documented proliferation of successful deployments and the business benefits realized by Network Equipment Providers and operators, the AdvancedTCA ecosystem is now driving enhancements to the technology and extensions to the standard. This allows these benefits to be brought to additional applications outside of the Central Office and, in some cases, outside the telecom industry altogether.

It is important to note that this trend is not simply being driven by vendors. Increasingly, the problems confronting companies running services have nothing to do with server capability. Rather, these problems most often stem from power consumption, cooling requirements, serviceability, and increasingly the cost of downtime. Designed for use in high-reliability telecom network applications, AdvancedTCA brings values such as open standards, inherent reliability, and power efficiency to the already established benefits of bladed architectures such as scalability and serviceability. As a measure of how importantly this is viewed, the number and value of “server-class” opportunities for Emerson’s AdvancedTCA solutions have tripled since 2007.

Therefore, the AdvancedTCA ecosystem is placing a greater focus on incorporating server-class attributes into next-generation products, including integrated and certified enterprise-grade operating systems, virtualization-ready hardware, and diverse storage expansion options. For example, Emerson is the first AdvancedTCA system provider to be certified and supported by leading information infrastructure provider EMC, and there are several ecosystem partners offering in-shelf storage expansion options.

What’s more, the applications outside the Central Office – where AdvancedTCA technology is now being considered – usually have a more relaxed cooling environment. For instance, the ASHRAE standards applied to data centers indicate a controlled environment with a maximum ambient temperature of 35 ºC (the NEBS equivalent is 55 ºC). In a 35 ºC environment, for the same airflow, more power can therefore be dissipated or lower-temperature parts can be used. Consider that in a 55 ºC NEBS environment with a typical airflow of 40 CFM per slot, it is just about possible to cool a server blade equipped with dual 60 W NEBS-grade Intel processors. Within the same airflow, but with an ambient temperature of 35ºC, it is possible to cool the same blade but now equipped with 80 W enterprise-grade (cheaper) processors. This improved performance, coupled with the native attributes of AdvancedTCA (open standard, reliable, power efficiency), appeal to non-telecom applications emphasizing service availability and thermal performance.

Military and medical take note

AdvancedTCA technology stands front and center as the military makes packet-based protocols key to intelligent warfare programs. For example, AdvancedTCA has been proposed as the core embedded computing technology for the U.S. Navy’s Consolidated Afloat Networks and Enterprise Services (CANES) program, which aims to consolidate the shipboard network infrastructure into a standard off-the-shelf solution.

New deployments have also expanded into cutting-edge medical and scientific laboratory environments. For example, a recent medical symposium heard a proposal for a flexible AdvancedTCA-based system to support Multimodality Positron Emission Tomography (MPET), a medical imaging and diagnostic technique. Another prime example is in the scientific research field, where AdvancedTCA is also already in use in several lead projects such as NASA’s Jet Propulsion Laboratory Deep Space Network and the Tokamak Fusion Reactor in the Czech Republic.

These types of programs reflect an underlying trend, similar in some respects to the next generation all-IP network evolution within the telecom industry. High availability, scalability and manageability – the traits inherent to AdvancedTCA – are critical for the success of large-scale experiments. The adoption of increased bandwidth infrastructures has also enabled the AdvancedTCA ecosystem to showcase the standard’s ability to adapt to new challenges.

Early on, Emerson Network Power championed AdvancedTCA beyond telecom, having coined the term Commercial ATCA to describe a product strategy that adapts the AdvancedTCA specification to take it beyond the Central Office and into commercial applications, including media and content serving, network-centric warfare, and Web/data hosting. Commercial ATCA aims to maintain the key values of AdvancedTCA but omit some of the harder NEBS telecom-specific features that are not valued beyond the telecom Central Office.

Emerson Network Power’s first Commercial ATCA product, the Katana 2000 ATCA Bladed Server (Figure 1), consists of two rack-mount servers. Benefits it aims to deliver include power efficiency, reliability, serviceability, scalability, investment protection, and environmental (green) factors such as reusability of core components versus disposal of a complete server. The Katana 2000 Bladed Server employs a 1+1 server blade architecture, with two server blades each featuring dual 80 W “commercial grade” Intel Xeon E5520/E5540 processors based on the Nehalem architecture.

Figure 1: The Katana 2000 ATCA Bladed Server (click graphic to zoom by 1.3x)

AdvancedTCA Extensions

PICMG plans to propose formal improvements to the ATCA standard for these new application areas. The organization has established a subcommittee to develop the standard, named ATCA Extensions at press time.

The brief for ATCA Extensions is fairly wide-ranging, encompassing everything from simple changes to cost-optimizing existing blades for non-central office environments to supporting new higher compute density system configurations. Such configurations would include double-wide server blades that make use of bigger heatsinks to further increase CPU power, and back-to-back enclosures to take full advantage of typical data center rack depths. However, the principal of backwards and forwards compatibility remains a key driving principle, in order to protect existing investments and keep the ability to arrive at the best solution – with freedom to choose among many components in the ecosystem.

Network Equipment Providers and the entire AdvancedTCA ecosystem have invested a considerable amount of development time and money to arrive where we are today, and this investment is paying off as system integrators across diverse markets reap the benefits. The standard has successfully bridged the gap between the telecom Central Office and proliferation into a host of other server-grade applications.

Keeps on ticking: AdvancedTCA barge test

Those of you old enough remember the Timex commercials from way back will know that certain solutions must demonstrate their ability to continue – no matter what.

To verify the continual operation of onboard (on a ship, for example) equipment throughout a blast scenario, Northrop Grumman recently tested Emerson Network Power AdvancedTCA equipment. Afloat shock testing is better known in the defense industry as barge testing. The test involved placing 60 pounds of MX-1 plastic explosives 24 feet deep and 20 feet away from a barge containing the functioning equipment – and then detonating it (Figure 2).

Figure 2: A Northrop Grumman barge test of Emerson Network Power AdvancedTCA equipment showed no equipment failures or system resets. (click graphic to zoom by 1.8x)

Emerson supplied a 14-slot AdvancedTCA shelf outfitted with two single-slot combined system controller and switch blades and two processor blades, each featuring dual Intel Xeon LV processors. A second, smaller two-slot system, again containing two Intel-based processor blades, was also included in the testing. Both systems were running Red Hat Linux.

The barge test team mounted the AdvancedTCA systems shock-isolated Northrop racks, of the type used for a variety of programs, and secured them to the floor of the barge. An umbilical optical Ethernet link connected the systems back to the barge test control room. This enabled constant monitoring and communication during the test. A stream of PINGs and other mechanisms verified continual operation of the equipment.

Test equipment and cameras in the barge measure the movement of the racks during the testing. Multiple blasts were conducted over several days to account for different rack positions, equipment weightings, and physical orientation. Test results showed no equipment failures or system resets for any of the Emerson products on board the barge.

Acronyms

IP Multimedia Subsystem (IMS)

Call Session Control Function (CSCF)

Home Subscriber Server (HSS)

Home Location Register (HLR)

Serving GPRS Support Node (SGSN)

Radio Network Controller (RNC)

Base Station Controller (BSC)

Brian Carr is a Strategic Marketing Manager for the Embedded Computing business of Emerson Network Power.

Emerson Network Power Embedded Computing

www.emersonnetworkpower.com/embeddedcomputing

brian.carr@emerson.com