On the fast bit counting algorithms of large arrays

There are many applications where some property of an array of values has to be evaluated and some overall characteristic determined. As an example we can mention the calculation the sum of an array of bytes. Simple and naive approach is to iterate through the elements of the array one by one and add the values into some accumulator variable. Such approach is valid for simple computer architectures with single central processing unit (CPU) without any optimizations for speed. Recently the modern CPUs are usually multi-core and are equipped with several cache memories and are optimized for high performance. To construct fast algorithm for such a simple task as to count bits in an array, one has to consider and utilize all available resources. This contribution will present some techniques to construct fast algorithm which will count bits of large arrays.