Very cool work but since its easy to get memory-bandwidth bound with SQL I would wonder if it would make more sense to use GPUs with 10X the bandwidth. I know

Very cool work but since its easy to get memory-bandwidth bound with SQL I would wonder if it would make more sense to use GPUs with 10X the bandwidth. I know FPGAs will be getting HBM as well so this might help. Disclaimer: I work at MapD, a GPU database company. (http://www.mapd.com http://www.mapd.com)