Where is the bottleneck in AdvancedTCA solutions?

AdvancedTCA architecture has become a stream line option for variety of vendors and applications including LTE RAN, security, cloud services, policy access control, IPTV, IMS and other deep packet inspection.  Telco Systems has deployed its networking solution worldwide with variety of carriers and we knew that one of the toughest thing for application developer will be to handle the switching solution.

So we invested very early (in fact we were the first company to have ATCA based solution in the market in 2004) a lot of efforts to take our carrier expertise and to transport it to ATCA switching blade. However as switching blades with multi 40G capabilities are coming to the market.

It became obvious that the bottle neck has now moved to the CPU line blade, as they run in 10G rates. So for example ,in a 14 slot chassis,  Telco Systems 2 x HUB4 which has capability of 320G uplinks while backplane inter-connect with CPU blade that have maximum blade of 2x12x10G which result in only 240G. The meaning is 80GE are Blocked by CPU Blade .We expect this to be changed during 2012, as some vendors announced plans to release with new CPU blades with 2 ports of 40G to each blade. So while this will close the gap between the switch and the CPU blades ,we should expect it to turn around again in 2013 as 100G switch blades will start showing up in the market.

Related Info: What is AdvancedTCA and its deployment challenges

If you are ATCA developer, where would you want the bottle neck to be? (write to us in the comments)


Related posts