This is also known as non-hierarchical clustering. In partitional clustering, each object is divided into smaller non-overlapping subsets (partition) that have similar characteristics. An observation can belong to one, and only one, cluster.
Here is a classic example of a three-cluster partitioned model. However, you may wonder why some of the yellow observations end up in the yellow group cluster, rather than the red group. That is because k-means has determined what the linear boundaries are for each of the clusters (right diagram), and any point that falls within any particular decision boundary is defined by the boundaries of that cluster: