Abstract: This paper presents a new kind of algorithm（GroupDFS） for computing cover quotient cube.The algorithm first get a result set by grouping the original base table based on all dimension attributes，then generate the upper bound set of the result set，finally return the cover quotient cube of the original base table.GroupDFS integrates the advantages of 2^N-algorithm and DFS algorithm，it shortens the time needed to compute the cover quotient cube.Experimental results on the dataset weather show that the time needed for computing when GroupDFS-algorithm is adapted is only 45% of that when DFS algorithm is adapted.