Infineon: CAN FD success goes at the expense of FlexRay

February 25, 2015 // By Christoph Hammerschmidt
The faster version of the venerable CAN bus, CAN FD is currently taking off at several carmakers. Infineon's Thomas Böhm, Head of Body / Automotive, believes this could well go at the expense of FlexRay.

In an interview during the Embedded World trade fair, Böhm said that from Infineon's perspective, CAN FD is currently gaining traction. "Inquiries are increasing rapidly", he said. While initially the Flexible Data rate version of CAN was driven by General Motors in the first place, the European OEMs are about to jump onto the bandwagon. The FD version offers basically the same functionality as the standard version developed in 1987; however it offers a and bigger payload and higher bandwidth. With this properties, it competes against the FlexRay data bus, which is driven by Infineon's competitor NXP in the first place. "We believe that CAN FD will eat up FlexRay's market share", Böhm said. "It is well possible that FlexRay gets under fire from two sides - CAN FD and Ethernet. While CAN FD has a cost advantage, Ethernet offers higher bandwidth. Infineon is prepared for the trend towards CAN FD, Böhm said: "We have the suited SPCs and transceiver modules ready".

In the medium term, Infineon also will support the DIS-2015 specification though thus involves a partial redesign of the respective products. "This will be the spec that goes into the (series) vehicles," Böhm said.

Also with regards to automated driving and the electronic sensing and control equipment, Infineon believes to be well prepared. The company hopes on orders for its Aurix microprocessor family which is based on the proprietary TriCore architecture. According to Böhm, These microprocessors are already very successful in radar sensors, Advanced Driver Assistance Systems (ADAS) and all kinds of safety-relevant applications up to criticality level ASIL D such as brake and steering control units. "When it comes to autonomous driving, the challenges shift from signal processing to data fusion and functional safety", Böhm said, adding that such applications require far higher processing power and four to five bigger RAM capacity.

Which brings us back to CAN FD: "For data fusion platforms, CAN FD is a