Can FPGA vendors help ATCA's fortunes?

 I wanted to point out an Actel Corp./Pigeon Point Systems press release today for two reasons: First, it’s interesting to see an FPGA vendor move upstream in any vertical market to offer better subsystem support; and second, the suite of Advanced Telecommunications Computing Architecture standards, which include MicroTCA and PicoTCA as well as ATCA, needs a bit of promotion if it’s going to go anywhere.

First of all, the details – Pigeon Point has been a specialist in shelf management and Intelligent Platform Management since well before the company’s 2008 acquisition by Actel. On March 29, Pigeon Point introduced an IPM Controller board based on a standard ATCA architecture, as well as a carrier card based on ATCA’s Advanced Mezzanine Card implementation. Both IPM solutions utilize Actel’s SmartFusion FPGAs with ARM Cortex-M3 controllers.

Pigeon Point emphasized the flash programmability of SmartFusion, and the configurable IP available in the FPGAs, which are relevant factors for much more than just touting Actel’s capability. ATCA and its derivatives are relegated, at least for now, to a market primarily defined by in-house development. PICMG (PCI Industrial Computer Manufacturers Group), the organization responsible for ATCA development, was hoping for more of an open board/open backplane development environment for ATCA. But the standard needs more tools like the Pigeon Point programmability to get anywhere.

The problem ATCA faces today has little to do with inherent flaws in the standard, or in a lack of cards from independent board vendors. Full-sized ATCA and AMC cards are offered for a variety of control card, shelf management, line card, and power management functions. Rather, the problem lies in the continued flatline state of the telecom infrastructure industry (both wireline and wireless), and the consolidation of players in the field.

If we count router vendor Cisco Systems as a telecom backbone player, then we see a telecom equipment market comprised of Huawei, Cisco, Alcatel-Lucent, Ericsson, and a few specialty players like Ciena and Fujitsu. That’s not a lot to hang one’s hat on. The problem is not a continued presence in telecom by Compact PCI – instead, the problem is that the few OEMs still around aren’t spending much time looking at open backplanes.

The MicroTCA proponents were hoping for a dedicated mil-aero market, but most military contractors remain heavily committed to VME and VPX. In short, both commercial and mil-aero telecom needs to revive as an industry before the wisdom of open backplanes for broadband services becomes self-evident. Pigeon Point and Actel can’t jump-start this industry single-handedly, but configurable tools like Pigeon Point’s IPMC boards certainly can’t hurt.