Fig. 11From: Parallel acceleration of CPU and GPU range queries over large data setsSpeedups for all GPU methods vs. a the iterative CPU method using 16 cores and vs. b the reduction CPU method using 16 cores. The Zipf data set skews are appended (e.g., Zipf0 is the Zipf data set with a skew of 0). The horizontal dashed line indicates a speedup of 1 ×. All plots share the same legendBack to article page