Don't have Telegram yet? Try it now!
https://hgpu.org/?p=5813
Hybrid coherence for scalable multicore architectures