Researchers have developed a method to increase by a factor of five the computing power of a standard algorithm when performed in one type of standard chip, FPGA. The new method is both simple and smart, but the road to publication has been long.