代码

#pragma GCC optimize("Ofast")
#pragma GCC target("avx2,popcnt,bmi,bmi2")
void count_5d(int n, const int **x, unsigned *out) {
#define n 100000
#define P(k) {x[k][i], x[k][i+1], x[k][i+2], x[k][i+3]}
#define Q(k) (r[0][k]<x[0][j] & r[1][k]<x[1][j] & r[2][k]<x[2][j] & r[3][k]<x[3][j] & r[4][k]<x[4][j])
	for (int i=0; i<n; i+=4) { int r[5][4] = {P(0), P(1), P(2), P(3), P(4)}; for (int j=0; j<n; ++j) {
out[j] += Q(0) + Q(1) + Q(2) + Q(3);
}}
}

评测结果