Power draw with GPGPU is better per flop than normal CPUs, but not competitive with an FPGA tuned to a specific problem with the same equivalent throughout. I

Power draw with GPGPU is better per flop than normal CPUs, but not competitive with an FPGA tuned to a specific problem with the same equivalent throughout. I think FPGAs also win out in flexibility. GPUs don't integrate so well into e.g. the signal processing hardware for digital oscilloscopes (a common use for high end FPGAs) mostly because of latency and timing constraints.