Don't have Telegram yet? Try it now!
https://hgpu.org/?p=19304
Sparse matrix partitioning for optimizing SpMV on CPU-GPU heterogeneous platforms