When FPGAs are better at floating-point than microprocessors

It has been shown that FPGAs could outperform high-end microprocessors on floating-point compu-tations thanks to massive parallelism. However, most previous studies re-implement in the FPGA the operators present in a processor. This is a safe and relatively straightforward approach, but it doesn’t exploit the greater flexibility of the FPGA. This arti-cle is a survey of the many ways in which the FPGA implementation of a given floating-point compu-tation can be not only faster, but also more accu-rate than its microprocessor counterpart. Techniques studied here include custom precision, specific accu-mulator design, dedicated architectures for coarser operators which have to be implemented in software in processors, and others. A real-world biomedi-cal application illustrates these claims. This study also points to how current FPGA fabrics could be en-hanced for better floating-point support.