Don't have Telegram yet? Try it now!
https://hgpu.org/?p=11373
Accurate Cross-Architecture Performance Modeling for Sparse Matrix-Vector Multiplication (SpMV) on GPUs