
Hence we used an FPGA-based near data processor developed in the CS12, which has query processing capability as well as an attached data storage. We built COPRAO by extending Apache Calcite optimizer with new optimization rules and cost model to consider this near data processors query processing capability.