1send (1) 0 g6<1>F g1.12<0,1,0>F math inv scalar mlen 1 rlen 1 { align1 }; 2send (1) 0 g6.4<1>F g1.20<0,1,0>F math inv scalar mlen 1 rlen 1 { align1 }; 3add (8) g7<1>F g4<8,8,1>F -g3<8,8,1>F { align1 }; 4mul (1) g7<1>F g7<0,1,0>F g6<0,1,0>F { align1 }; 5mul (1) g7.4<1>F g7.4<0,1,0>F g6.4<0,1,0>F { align1 }; 6mov (8) m1<1>F g7<0,1,0>F { align1 }; 7mov (8) m2<1>F g7.4<0,1,0>F { align1 }; 8mov (8) m3<1>F g3<8,8,1>F { align1 }; 9send (8) 0 null g0<8,8,1>F urb 0 transpose used complete mlen 4 rlen 0 { align1 EOT }; 10nop; 11nop; 12nop; 13nop; 14nop; 15nop; 16nop; 17nop; 18