[dheera]

Why not just

   (a >> 1) + (b >> 1) + (a & 1) & (b & 1)

[dreamcompiler]

That's the "patented solution" referred to in the article.

[andi999]

Which is shocking to me, since this is the solution which jumps immediately to mind (and usually patents should be above that line). I could understand if the next solution was patented, but not this one.

[dheera]

Don't read the article, and you would never know it's patented.

[Dylan16807]

"just"? That's 2 additions, 2 shifts, and 2-3 bitwise operations, while the other method is 1 addition, 1 shift, and 2 bitwise operations.

[rustybolt]

I would think optimizing compilers can optimize this expression to the one in the article, but I tried it on godbolt and they don't.