A Better Parallel Scan Algorithm 1.Read input from device to shared memory 2.Iterate log(n)times;stride from 1 to n-1:double stride each iteration 3.Write output from shared memory to device memory XY 3 4 3 STRIDE 1 XY 3 8745 STRIDE 2 XY 3 4111112121114 STRIDE 4 个 XY 3 4 1111151622 25 ITERATION 3 STRIDE =4 电子科妓女学 O15 A Better Parallel Scan Algorithm 1. Read input from device to shared memory 2. Iterate log(n) times; stride from 1 to n-1: double stride each iteration 3. Write output from shared memory to device memory XY 3 4 8 7 4 5 7 9 XY 3 1 7 0 4 1 6 3 ITERATION = 3 STRIDE = 4 STRIDE 1 XY 3 4 11 11 12 12 11 14 STRIDE 2 XY 3 4 11 11 15 16 22 25 STRIDE 4