Open access
Aug 2026
Structure-Derived Bottleneck-Aware Scheduling for Multitasking MCM-GPUs
SA-Scheduler is presented, a structure-derived bottleneck-aware scheduling framework for multitasking MCM-GPUs that determines chip placement without hardware modification or runtime bottleneck profiling, and provides a principled and scalable foundation for multitasking on future MCM-GPUs.
Tiejian Zhang, Guangda Zhang, Lu Wang et al.
· ACM Transactions on Design A... · 0 citations