Don't have Telegram yet? Try it now!
https://hgpu.org/?p=16643
Reordering strategy for blocking optimization in sparse linear solvers