Time to get religion about God boxes

Religion remains a controversial topic, even in the world of telecommunications networking equipment. The controversy centers on a type of equipment called "God boxes."

The term "God box" has been around for several years and generally refers to telecom equipment that can perform a long list of functions in wireline and wireless equipment. Specifically, a God box consists of a chassis (equipment box) housing a plethora of network system functions, such as optical switching, core routing, Asynchronous Transfer Mode switching, add drop multiplexing and dense wavelength division multiplexing. Put another way, a God box consolidates several different telecom equipment systems into one and provides a full range of service and capacity requirements. Such a box amounts to a communications bridge that spans from telecom transport, using currently installed Synchronous Optical Network technology, to Ethernet links, to next-generation packetized Multi-Protocol Layer Switching IP services.

One vision is for the industry to continue moving toward integration of  access services and functions traditionally in digital loop carrier, access switching and transport and access data systems into a Digital Subscriber Line Access Multiplexer (DSLAM) or broadband access node. This type of integration of different boxes into other boxes, to consolidate systems and reduce operational and capital expenses, is bound to continue.

Sounds good so far. But here's the catch: Despite their obvious benefits, including reducing the amount of equipment required to handle telecom traffic and thereby lowering equipment costs, until recently God boxes have had the dubious reputation until of not doing any of the functions they are supposed to especially well. And they have been judged to be uneconomical to build and use.

But such God boxes in the access network are gaining a more positive reputation. Such equipment has recently been introduced by a small but growing number of system vendors in the United States and abroad. As it turns out, God boxes can perform better and be made more affordable than skeptics predicted.

Indeed, consolidating several different equipment platforms into one has proven to be achievable and a winning business initiative. One type of God box emerging in prominence is the Multiservice Access Node (MSAN). Located in the access portion of the telecom network, MSANs consolidate DSL, broadband and phone services from neighborhoods and homes, as well as high-speed voice and data services for business. 

One reason God boxes are materializing is ongoing development and delivery to market of innovative technologies. They include resiliency-on-demand and robust traffic management quality of service capabilities. Resiliency-on-demand equips a DSLAM with the flexibility to guarantee a service will not fail for the service provider's highest-paying customers. 

For example, a wireline service provider will guarantee the bandwidth and visual or audible quality of a service, such as a voice call or video. The provider will also offer an uninterrupted service experience. Achieving economical service resilience in broadband means the service provider will only guarantee delivery of services its customers want. Such customers won't have to pay to guarantee fast delivery of other lower-priority services. Such guaranteed services include voice over Internet protocol, video streaming/conferencing, and real-time interactive online gaming. Network delays and service interruption for such applications are not tolerated and greatly degrade the subscriber experience.

The kinds of semiconductor chips that will benefit from this God box trend include those that are used in DSLAMS, such as network processors that perform robust traffic management functions. A God box must help a telecom service provider generate more revenue than they can with current equipment. Traffic management enables a service provider to offer per-service guarantees of services to its highest paying customers and reduce the amount of services it guarantees, which reduces its operational costs.

Other types of chips that stand to benefit include chips that integrate mapping, framing, traffic management, Ethernet and Synchronous Optical Network functions. For example, such single chip solution eliminates the need for a separate equipment box that performs SONET but not Ethernet, or framing but not mapping; or traffic management but not SONET. This reduction in equipment boxes translates to lower overall system, silicon and electronics costs.

With the growing use of these God box systems, the overall system market is likely to decrease in the number of systems built because more of them will be consolidated into a fewer overall number. The market for more versatile boxes that can do more functions is likely to grow. The market for more integrated chips that can handle more major functions such as SONET, Ethernet and traffic management is likely to expand as the market for single-function chips is bound to decline.

God boxes will likely take away from use of other system approaches that integrate fewer functions because the latter approaches will tend to cost more in equipment and chip costs. For example, a system box that integrates traffic management and Ethernet functions but not SONET and mapper functions will likely decline in use and be displaced by boxes that integrate all these functions.

There will be some chip interface standards that chip makers will need to design to get in on the God box business. But such interface adherence won't be a major issue. Chips will have to adhere to Ethernet and SONET standards, for example, but chip makers have to make them adhere to such standards even if such chips are housed in a several different boxes rather than a God box. The interface issues won't be a significant factor in accelerating or slowing this market.

God boxes will continue to be a significant factor in accelerating the integration of various different types of semiconductor chips within one chip and one system. Put another way, they will accelerate the system-on-a-chip trend. And such boxes will be equipped with more revenue-generating and cost-reduction features such as quality of service and box size reduction.  So God boxes will help accelerate the amount of revenues wireline and wireless service providers can generate from their customers. With increased revenues they will be able to pay to upgrade their networks with more God box equipment, which they will increasingly need to remain competitive in the market.

In summary, ongoing advances in telecom network resiliency and traffic management will be key to spreading the God box and God box chip integration religion, which is already attracting a fast-growing number of believers.

Chris Hamilton is the senior manager of systems architecture at Agere Systems.

Visit Agere Systems online.