September 2016
Intermediate to advanced
774 pages
24h 26m
English
B. Wu1; X. Shen2 1 Colorado School of Mines, Golden, CO, United States2 North Carolina State University, Raleigh, NC, United States
Task scheduling is critical for exploiting the full power of multicore processors such as graphics processing units (GPUs). Unlike central processing units, GPUs do not provide necessary APIs for the programmers or compilers to use to control scheduling. On the contrary, the hardware scheduler on modern GPUs makes flexible task scheduling difficult. This chapter presents a compiler and runtime framework with the capability to automatically transform and optimize GPU programs to enable controllable task scheduling to the streaming multiprocessors (SMs). ...
Read now
Unlock full access