At the Advanced TCA Summit industry forum in San Jose California, I had the opportunity to discuss my views on the evolution and future of ATCA. The specific topics were, “ Why was ATCA successful” and “what will the future hold?”. This was part of a session recognizing key contributors to ATCA, and I was very proud to be recognized as a key contributor over the past 10 years.
I believe ATCA is the most successful building practice ever developed in telecom. It is approaching $1B in sales, 10 of the top 10 telecom equipment manufacturers use the technology, hundreds of companies are developing products built around ATCA, and it has evolved through three generations of Ethernet advances (1Gigabit to 10G to 40G) over the course of 10 years.
ATCA was successful for several reasons, but primarily because it had excellent technology, and a very strong business value proposition, and was pushed aggressively by a large company and many smaller ones. ATCA started with a strong specification from a strong and respected standards body. ATCA was originally specified in 2002 led by many members of the PICMG, an industry association. (http://picmg.org/v2internal/specifications2.cfm?thetype=One&thebusid=2). The technology addressed known issues with previous standards and was widely available, with good IP protections. PICMG met the test for credibility, openness, and competence. Many technologists think this was enough; PICMG was necessary , but not sufficient to the success of ATCA.
Great standards must have great commercial success in the market. ATCA has enjoyed great commercial success and there were several reasons for this.
First, ATCA was a compelling value as a common platform for many network elements. ATCA came at the right time for a common platform as telecom was hit hard by the slump of 2001-2005, and new platforms had to be developed with reuse in mind. Chuck Byers of Cisco has shown up to 20% savings with a good common platform approach. For a large company looking to cut costs without cutting programs, this was very important.
Second, Intel was behind ATCA for several years and had the access to key decision makers to help them buy into the technology. Intel needed ATCA to provide a suitable power budget for both x86 and network processors, and was keen to provide boards and systems at the time. (Previous telecom standards did not allow for the most powerful and most expensive Intel processors). Intel invested heavily and delivered the first systems while enabling other companies with products. Intel also lined up several smaller companies who were able to move very quickly to complement Intel’s products, and provide complete solutions to customers. These included Force Computer, Znyx, Go Ahead, Monte Vista and others. Three of these four companies were acquired later on for good valuations, in part because of the leadership position they established in ATCA.
Finally, ATCA attracted several big design wins that established momentum worldwide. I do not think you can overstate the importance of several ground breaking design wins in North America, China, Japan, Korea, and Europe. These wins came about because key decision makers were sold on the technology and were willing to bet their next designs, and in some cases their careers on ATCA. There were internal advocates within many of these companies, but also a big push from Intel and the ecosystem. (A side note, many of the early adopters of ATCA at leading telecom companies have been promoted 2 or more times as a result. Fortune favors the bold.....)
This leads to my conclusions on what a successful industry initiative needs:
ATCA had all of these in 2002-2005 and we are seeing the results today.
Next Topic: What is the future for ATCA?