import jdk.incubator.vector.*;

static int countDifferingBits_1024bit_vectorAPI(long[] array1, long[] array2, int i1) {
  VectorSpecies<Long> SPECIES = LongVector.SPECIES_PREFERRED;
  
  int diff = 0;
  int n = 1024/64;
  var upperBound = SPECIES.loopBound(n);

  var i = 0;
  for (; i < upperBound; i += SPECIES.length()) {
    var vector1 = LongVector.fromArray(SPECIES, array1, i);
    var vector2 = LongVector.fromArray(SPECIES, array2, i1+i);
    var xored = vector1.lanewise(VectorOperators.XOR, vector2);
    var bitCount = xored.lanewise(VectorOperators.BIT_COUNT);
    diff += (int) bitCount.reduceLanes(VectorOperators.ADD);
  } 
  
  // Compute elements not fitting in the vector alignment.
  for (; i < n; i++)
    diff += Long.bitCount(array1[i]^array2[i1+i]);

  ret diff;
}

Began life as a copy of #1036056