Comparing LLC-Memory Traffic between CPU and GPU Architectures
Event Type
Workshop
Architectures
Extreme Scale Comptuing
Heterogeneous Systems
W
TimeFriday, 19 November 202111:23am - 11:26am CST
Location229
DescriptionThe cache hierarchy in modern CPUs and GPUs is becoming complex. The introduction of additional complexities by manufacturers makes understanding the handshake between the memory access pattern and the cache hierarchy difficult. Moreover, the details of different cache policies are not publicly available. Therefore, the research community relies on observation to understand the relation between memory access patterns and cache hierarchy. Our previous studies delved into different micro-architectures of Intel CPUs. In this study, GPUs from NVIDIA and AMD are taken into consideration. Even though the execution models in CPUs and GPUs are distinct, this study attempts to correlate the behavior of the cache hierarchy of CPUs and GPUs. Using the knowledge gathered from Intel CPUs, the similarities and dissimilarities between CPUs and GPUs are identified. Through model evaluation, this study provides a proof of concept that LLC-memory traffic can be predicted for sequential streaming and strided access patterns on GPUs.
