1draw1: 6M for draw 0,0,100,100 no repl 2draw3: 4M for draw 0,0,100,100 no repl 3just read src, dst - 250k 4mask reading - 650k 5write dst - 100k 6alpha calculation - 3000k 7 8olddraw: 10M for draw 0, 0, 1000, 1000 no repl all ldepth 3 9 44M for draw 0, 0, 1000, 1000 src, mask ldepth 2 dst ldepth 3 10draw4: 160M for draw 0, 0, 1000, 1000 no repl all r8g8b8 11 null loop: 10k 12 src, dst reading: 13-15M each 13 mask reading: 30M 14 alpha calculation loop: 90M 15 null alpha loop: 2M 16 minimal loop control +20M 17 alpha calculation with divides +190M 18 alpha calculation wtih shifts +70M 19 writeback: 11M 20