ATCA Can Add Value to the Telephony Market. . . But When and How?

Since the base specification release of PICMG 3.0 R1.0, AdvancedTCA (News - Alert) Specification for Next Generation Telecommunications Equipment, the increasing level of interest in this design architecture has increased each and every year. After the initial release of the standard at the end of 2002, the growth projections by the year 2007 were staggering, but thus far the actual realized revenue growth has not lived up to those expectations. My objective for this column is to provide readers with a “street view” of what’s happening with the technology.

So let’s explore the state of the AdvancedTCA’s (ATCA) progress and see who is winning the game. Since I am a sports enthusiast, I like to break things up into logical segments — like a game plan. New technology adoption typically flows in four quarters:

1. Technology evaluation — “Does it really work and how?”
2. Acceptance — “What’s the value and can we make money?”
3. Deployment — “Production and lower cost activities.”
4. Phase out — “Final days.”

Currently the ATCA products are stuck in halftime. Many hardware and software vendors are still trying to justify and re-coop their significant capital investments in the technology. At the same time Independent Software Vendors (ISVs) and Original Equipment Manufacturers (OEMs) are trying to determine how ATCA hardware will enhance their solutions. Both seemingly are stuck half way.

One way to get a fast read on the state of the technology is to ask the creators and users some probing questions. So I quickly tapped into a network of experts that have firsthand experience with the good and the bad when it comes to ATCA.

What are the most important recent developments in ATCA technology?
The consensus among the network of experts is that the ATCA base standard is good, but its offspring are really the market-makers. Follow-on standards such as the AMC PICMG specifications (.0, .1, .2, and .3) and the MicroTCA (News - Alert).0 R1.0, based upon AMC cards that plug directly into a backplane, seem to be where the action is.

According to Rubin Dhillon, GE Fanuc Global Director of Embedded Communications Products, “The most important recent developments in AdvancedTCA are the great strides the AdvancedMC (News - Alert) specification has taken — specifically AMC.0 ECR001, an Engineering Change Request to add additional, new AMC connector types and vendors to the AMC.0 specification, and AMC.0 ECR002, a more general Engineering Change Request to review and update the AMC.0 specification, add features and correct errors in the specification. These changes will enable AdvancedMCs to give ATCA designers more flexibility and add a wider range of interfaces to each design, meaning that identical base systems can be customized with different processing and I/O capabilities for specific applications. The original goal for AdvancedMCs was to define an ATCA mezzanine card that would meet the modularity and scalability requirements of the telecom industry. Its focus is therefore high availability, manageability, hot-swappability and a serial switched fabric backplane. Ultimately, ATCA blades equipped with hot swappable AdvancedMCs provide the greatest modularity of any open telecom platform. Further, AdvancedMCs extend ATCA’s high performance switched fabric by providing data rates up to 200 Gbps. The benefits of modular computing are driving a general trend towards modularity, particularly within industries where major system development continues under pressure to cut costs and increase performance.”

Amnon Gavish, VP of Business Development for Radvision (news - alert) adds: “The current processors on the ATCA blades are not sufficient to handle all of our video processing, so our video transcoding units will require MicroTCA to host DSP-based Advanced Mezzanine Cards (AMCs) to do the work. This will enable us to perform the video transcoding between video and audio codecs as well as video enhancements such as picture in picture and text overlay.”

What hurdles does the ATCA standard still need to overcome?
Since the initial release of the ATCA base standard in December 2002, we have seen two major updates to the standard adopted and Release 3.0 is currently under development, but there is still work to be done.

“Many of the key attributes of the Advanced TCA architecture are well suited to a wide range of applications across the enterprise space as well as the telco space. This would include applications such as network security, VoIP, Voice Quality, etc. However — on the enterprise side, comes an increased focus on price /performance — and the need for a set of products which leverage the ATCA architecture — but with a subset of the high availability, NEBS/Telco features. These requirements are not specifically addressed in the ATCA specification — but clearly there is a market need for products in what can be dubbed the ‘ATCA-ish’ category — particularly from the chassis and management perspective. Ongoing development — particularly in the management area will be needed to facilitate standardization in the approaches.” — Laurie Burger, Business Development Director Pentair/Schroff.

Additionally, there has been a strong interest from the Military, Aerospace, and Government sectors in both ATCA and MicroTCA technology. This demand is fueling design requirements for ruggedized ATCA and MicroTCA-based platforms, but to date there is no official standard to specify temperature, shock, and vibration requirements.

“Beginning in 2004, we entered the ATCA marketplace and had some almost immediate successes in NEBS high density compute environments. Essentially, we were replacing a rack full of NEBS 2U servers with ATCA blades. This area of the market, however, has been limited to a few key customers. PICMG 3.0’s base connectivity at 1Gbps was just too slow for what many of our customers needed and a lack of off-the-shelf DSP / Telco network connectivity blades essentially stifled the tier 2 and 3 markets. . . Now, with PICMG 3.1 in prime time offering greater network performance, plus manufacturers like Dialogic, AudioCodes, Surf Communications (News - Alert), and others releasing their network building blocks in ATCA and AMC form factors, we are beginning to see a resurgence of interest in ATCA design work. Based on our current design work and customer trials, we expect our ATCA revenues to nearly double over 2006 and continue this growth into 2008.” — Austin Hipes, Director of Technology, Alliance Systems (News - Alert).

What ATCA architecture advantages allow you to differentiate your solution?
The key ATCA architectural advantages that seem to climb to the top of the list are:

1. Interconnect Bandwidth
2. Scalability
3. Density

First the interconnect bandwidth. The older CompactPCI (News - Alert) framework is limited to 1 Gbps per link under PICMG 2.16 and 4 Gbps for PCI. The ATCA standard blades will support up to 10 Gbps bandwidth per link. RADVISION, headquartered in Fair Lawn, New Jersey, is a leading provider of products and technologies for unified visual communications over IP and 3G networks. According to Amnon Gavish, VP of Business Development for RADVISION, “At RADVISION, we have a video web conferencing solution with a soft MCU which is a very demanding video application called Click to Meet ® Server. We were able to port our application over from CompactPCI to ATCA in a just a couple of weeks. The ATCA architecture gave us a significant increase in throughput both in and out of the blades and allowed us to triple the number of concurrent users on each blade.”

Secondly, the scalability of the ATCA architecture allows ISVs and OEMs to deploy a system which is populated with the right number of application blades to meet the user’s initial requirements. As the end user’s needs grow, they can add application blades to the existing environment without interoperability concerns. The end user can significantly lower both their CAPEX and OPEX (News - Alert) spending with this “grow with demand” strategy.

Lastly, the ATCA board surface area is more than 140% greater than a standard CompactPCI board. This increased board real estate allows hardware vendors to increase the board density inside the ATCA chassis. If we consider a PICMG 2.16 cPCI high density system to an ATCA high density system, the total revenue-producing area is nearly double (see table).

Final score
The field of experts agrees on two things. One, the adoption rate has been slower than expected, but momentum is building and the ATCA standard is emerging from halftime. The winners in the second half will be those who reap the benefits from AMC flexible designs, 10 Gbps throughput, and price advantages of MicroTCA.

Jeff Hudgins is VP of Engineering at Alliance Systems. For more information, visit the company online at http://www.alliancesystems.com.