K-means is a versatile and popular clustering algorithm with appli- cations in numerous scientific and engineering areas, such as com- putational biology, economics, and machine learning. One drawback of K-means is its inability to identify non-linearly separable clusters, which may lead to inaccurate solutions in certain cases. Kernel K- means is a variant of classical K-means that can find non-linearly separable clusters. However, it scales quadratically with respect to the size of the dataset, taking several minutes to cluster even medium-sized datasets on traditional CPU-based machines. In this paper, we present a formulation of Kernel K-means us- ing sparse-dense matrix multiplication (SpMM) and sparse matrix- vector multiplication (SpMV), and we show that our formulation enables the rapid implementation of a performant GPU-based ver- sion of Kernel K-means with minimal programming effort. Our implementation, named Popcorn, is the first open-source GPU- based implementation of Kernel K-means. Popcorn achieves a speedup of up to 123.8× over a CPU imple- mentation of Kernel K-means and a speedup of up to 2.6× over a GPU implementation of Kernel K-means that does not use sparse matrix computations. Our results support the effectiveness of sparse matrices as tools for efficient parallel programming.