@Dylan McGrath: yes, Flipchip has been growing for years now. The new growth is also coming from non-solder based bumps (conductive adhesives) which can be placed not just on substrates but also on Si interposers which are now growing in 2.5D integration. The Si interposers are taking some of the limitations on number of I/O's away because of their accommodation of less than 3mil lines and pitch.
@chipmonk: to partially answer your question as to why more ASAT's are not getting in to the act -basically it is an IP issue. There are two major IP holders for bumping technology and the associated underbump metallurgy (UBM) patents, which are Unitive (bought out by Amkor) and FlipChip Technologies.
Other reasons could be the development of processes on a large volume scale. Most are bonded thermosonically but the process parameters change significantly for solders, conductive adhesives, microbumpts of different metals and Cu pillars.
Dr. MP Divakar
Current growth of flip chip is linked to the growth in portable electronics e,g. smart phones.
Profit margin for flip chip is still higher than older packaging technologies. Even Cu uPillar bump technology which is being touted by the established OSATs ( with operations in the Far East ) as leading edge and being marketed at a premium has actually been in production in one form or another in the US for nearly 15 years and the technology is stable & easily transferable. Capital requirements for wafer bumping & flip chip / WLP are relatively modest ( $ 100 million ).
So why are more OSATs ( e,g. from China ) NOT getting into the act ?
Blog Doing Math in FPGAs Tom Burke 16 comments For a recent project, I explored doing "real" (that is, non-integer) math on a Spartan 3 FPGA. FPGAs, by their nature, do integer math. That is, there's no floating-point ...