A Survey on Recent Hardware and Software-Level Cache Management Techniques

Abstract

Fig. 1. Slowdown of bzip co-located with different applicationsAbstract-Multi and many-core processors have emerged as the dominant solution for processing in the whole range of computer system, from small devices to large-scale installations. Chip multi-processors, which are homogeneous, multi and manycore processors, offer an unprecedented amount of on-chip, shared resources and brings a unique set of challenges. Given the importance of the Last-Level Cache management techniques to achieve near-perfect isolation, we survey the state of the art and propose research directions to address the most pressing issues in modern computer systems. To better understand the various research directions in the field, we propose a classification of the presented techniques. Finally, we discuss possible research directions.

Multi-core Devices for Safety-critical Systems

“…Cache components [149,150,151,152] reduce memory access time of computing cores, bridging the frequency gap between cores that operate at higher frequencies than memory components. A multi-core device might integrate private caches assigned exclusively to given cores (e.g., L1 cache) and public caches shared by several components such as cores (e.g., L2 cache).…”

On-Chip Dynamic Resource Management

“…A number of surveys offer systematic overviews of parts of the topics, challenges and techniques that we are concerned with. Some of them focus on specific subsystem of a many-core chip like Networkson-Chip (NoC) (Bjerregaard and Mahadevan, 2006;Marculescu et al, 2009;Rahmani et al, 2010;Radetzki et al, 2013) or caches (Scolari et al, 2014), others deal systematically with particular metrics like power, energy (Maiterth et al, 2018), or aging (Khoshavi et al, 2017). Also, scheduling, mapping and task allocation have received a fair amount of attention in surveys (Singh et al, 2013a(Singh et al, , 2017Zhuravlev et al, 2012;Burns and Davis, 2017), with focus either on performance Introduction approach contributes to multiple metrics (e.g., energy, temperature, QoS, etc.)…”