FPGA based acceleration to the rescue from the death of Moore's Law

Great article:

They cited the use of more cores as a means for the semiconductor industry to work around the problem, however they totally missed the under tide of trends taking place to hardware accelerate functionality with FPGA fabric coupled to the processor. Until real alternatives arrive in the next decade or more, I think we're going to see more FPGA based optimizations taking place as an extension of the processor along side multi-core parallelization. I'm more optimistic about security, and network function acceleration in the fabric than I am on the conversion of legacy applications to use multiple cores. The lower the optimizations the more vast their impact so it makes sense for investing time and energy on accelerating these layers for the cloud and big data market segments.