This paper presents Global Density Clustering, (GDC), an algorithm that has several major advantages over the most popular existing clustering algorithms: (1) No parameters are chosen at the outset of the function; rather, the user can control the desired resolution as clustering proceeds. (2) GDC is efficient enough to work on a large dataset even when there are a sizable number of features. It is O(MN log N) where M is the number of features, i.e. the dimension, and N is the number of data points, i.e. the dataset size. It is suitable for big data. (3) GDC has the advantage of the powerful and intuitive definition of clusters as: points within a cluster are closer than distance dist to their nearest neighbor in the cluster (dist is not picked at the outset but rather that is chosen as the algorithm is progressing) and all points outside the cluster are further than dist from any point in the cluster. (4) GDC supports variable density without the plethora of special data structures such as HDBSCAN needs. (5) Other advantages are described. An essential reason that GDC has these advantages is it searches for and considers points whose nearest neighbor are furthest apart before searching for those that are closer together. It is a top-down or “global” consideration of distances that other density algorithms do from a bottom-up or “local” view. Other novel approaches to the main problems of clustering such as noisy backgrounds are described.