[en] The current status of lattice Quantum Chromo Dynamics (QCD) calculations, the computer requirements to obtain physical results and the direction computing is taking are described. First of all, there is a lot of evidence that QCD is the correct theory of strong interactions. Since it is an asymptotically free theory we can use perturbation theory to solve it in the regime of very hard collisions. However even in the case of very hard parton collisions the end-results of the collisions are bound states of quarks and perturbation theory is not sufficient to calculate these final stages. The way to solve the theory in this regime was opened by Wilson. He contemplated replacing the space-time continuum by a discrete lattice, with a lattice spacing a. Continuum physics is then recovered in the limit where the correlation length of the theory, say ξ. is large with respect to the lattice spacing. This will be true if the lattice spacing becomes very small, which for asymptotically free theories also implies that the coupling g becomes small. The lattice approach to QCD is in many respects analogous to the use of finite element methods to solve classical field theories. These finite element methods are easy to apply in 2-dimensional simulations but are computationally demanding in the 3-dimensional case. Therefore it is not unexpected that the 4-dimensional simulations needed for lattice gauge theories have led to an explosion in demand for computing power by theorists. (author)