Donghyeon Joo

Characterizing LLM Kernel Access and Memory Interaction in Multi-Partition NUMA GPUs

This work categorizes LLM kernel operands into three inter-workgroup sharing patterns and shows that the required optimization strategies differ across categories, from simple per-workgroup pinning to subgroup-aware co-scheduling, highlighting the need for placement-aware kernel programming and smarter architectural support for work and data locality in multi-partition GPUs.