Is available memory taken into consideration when scheduling pods?

Just found this post that is asking the same question. Not sure why I missed it when searching earlier