The need for fast hyperspectral data processing methods is discussed. Discussion includes the necessity of faster processing techniques in order to realize emerging markets for hyperspectral data. Several standard hyperspectral image processing methods are presented, including maximum likelihood classification, principal components analysis, and canonical analysis. Modifications of those methods are presented that are computationally more efficient than standard techniques. Recent technological developments enabling hardware acceleration of hyperspectral data processing methods are also presented as well as their applicability to various hyperspectral data processing algorithms.