Most error estimates of numerical schemes are derived in the field of real or complex numbers. From a computational point of view this assumes infinite precision. For the implementation on a computer, the infinite number fields are quantized into a finite set of values. Numerical stability analysis of the schemes then reveals how sensitive they react to distortions of the data, introduced by the data quantization and rounding. However, precise quantitative analysis of complex schemes with a clear dependence on the quantization parameters is very difficult. In particular, for iterative schemes which feed the results of one iterative step as input into the next step, and possibly execute hundreds or even thousands of iterations, there is no easy way to obtain reliable error bounds in the long run. This talk is going to give an introduction to mixed precision methods which helps in obtaining the same precision for less computational resources and power.