CoAdELL: Adaptivity and Compression for Improving Sparse Matrix-Vector Multiplication on GPUs | IEEE Conference Publication | IEEE Xplore