AMD has disappointed industry watchers by suggesting that its next-generation Carrizo APU might be 28nm instead of the hoped for 20nm.
AMD was thought to have the time and tech to go for 20nm with its foundry partners, but LinkedIn data suggesting that Carrizo, AMD’s first APU with an Excavator-class CPU, is built on the much less interesting 28nm.
This suggests that AMD is waiting to move to 20nm until that node is more mature, at least for big-core CPUs. Instead it is using GF28A, s GlobalFoundries (GloFo) standard bulk silicon node. Profiles from current and former AMD employees point to Kaveri as being built on 28nm SHP (Super High Performance) which is another big yawn.
According to ExtremeTech it is not clear if Kaveri and Carrizo are built on two fundamentally different types of 28nm silicon, or if the different codenames reflect subtle changes. AMD has been boasting of a 65W target for Carrizo and if that is the case then it is possible that AMD is moving to a different node that emphasizes lower power and higher efficiency.
AMD leaked another variant around Christmas which showed that the new Toronto APUs based on Excavator cores and DDR3/DDR4 controllers on the map for 2015, with and without a functional on-board APU. This was bad news for those hoping that AMD would return to shipping 8-core parts, and means that AMD is finished with larger CPU configurations.