I am often asked, "What are the most significant recent advancements in AdvancedTCA (ATCA)?" and I categorize them into 3 areas: 10G platform deployments, a new 40G backplane, and server applications.
First, there were a number of significant wins and the first real deployments of 10G platforms, which in turn created demand for the packet processing blades and systems to make use of this new fabric capacity. As predicted, this led to a big uptick in interest in data and packet-intensive applications. From an application perspective, this aligned well with the need to develop several new network elements for WiMAX and LTE wireless infrastructure. Generally speaking, as the network capacity increases users simply find more ways to fill the bandwidth and this drives the need for service platforms, packet management devices and access technology improvements, as well as billing engines and video infrastructure ... and, well I could go on. These are all applications where ATCA can and will play.
Equipment providers are increasingly interested in ATCA as an open industry-standards means of creating new platforms that can apply to many different new applications. This is especially true as the technology advances in say packet processing and internal fabric bandwidth increases the relevance of the technology. Service providers are also interested because they recognize that basing their infrastructure on an open standards platform will reduce their exposure to lock-in and will help their equipment suppliers advance their technology at more reasonable costs than proprietary systems.
Second, from an enabling technology perspective, there is no doubt that the advent of the 40G-KR serial backplane was the year's most exciting development in ATCA. Defined by the IEEE, this new backplane allows ATCA to offer up to 40Gbit/s Ethernet bandwidth per blade. As ATCA becomes increasingly applied to data plane applications such as packet flow processing and routing, the most important aspect will be installing systems that are ready for upgrade to 40G per blade. This is not because hardware will suddenly exist to fill and process many 40G pipes, but it is simply the case that the next technological advance in blades will be to go beyond 10G. In order to avoid a future "forklift upgrade" by changing out the backplane to support next generation blades, it is important to customers to install 40G ready systems now.
And finally is the trend towards using ATCA for server applications where equipment makers can use the common platform infrastructure they have already developed and apply it to telecom server applications such as customer databases.
A key technology enabler in the ATCA market is the just announced Intel® Xeon® 5500 Platform based on the latest-generation Intel® Microarchitecture (Nehalem). This architecture potentially offers a significant power advantage with all the new virtualization and power management features at a time when power/performance ratio will become a key metric.
On a concluding note, I'll mention that there is growing interest in using ATCA outside the telecom central office for which it was originally intended. I'll write more about the concept of "Commercial ATCA" in my next post.
All told, ATCA has made some significant strides. What progress do you think we should expect this coming year?