A word RAM is a unit-cost random-access machine with a word length of w bits, for some w, and with an instruction repertoire similar to that found in present-day computers. The simple lower bounds for the problems of sorting and searching valid in the comparison-based model do not hold for the word RAM, so that the well-known algorithms for these tasks are not optimal for the word RAM. This paper gives an overview of faster algorithms known for sorting and searching on the word RAM, many of which were developed within the last few years.