Question

Update: Please read the code, it is NOT about counting bits in one int

Is it possible to improve performance of the following code with some clever assembler?

uint bit_counter[64];

void Count(uint64 bits) {
  bit_counter[0] += (bits >> 0) & 1;
  bit_counter[1] += (bits >> 1) & 1;
  // ..
  bit_counter[63] += (bits >> 63) & 1;
}

Count is in the inner-most loop of my algorithm.

Update:
Architecture: x86-64, Sandy Bridge, so SSE4.2, AVX1 and older tech can be used, but not AVX2 or BMI1/2.

bits variable has almost random bits (close to half zeros and half ones)

Answer 1

Maybe you can do 8 at once, by taking 8 bits spaced 8 apart and keeping 8 uint64’s for the counts. That’s only 1 byte per single counter though, so you can only accumulate 255 invocations of count before you’d have to unpack those uint64’s.