Paper Info
Title
STLAC: A spatial and temporal locality-aware cache and network-on-chip codesign for tiled many-core systems
Abstract
The spatial and temporal locality of workloads are the root causes for cache designs to overcome the memory wall problem. However, few existing state-of-the-art designs exploit both the two locality features to optimize the memory hierarchies in the area of tiled many-core systems, which losses the opportunities to explore more performance improvement. To address this problem, an adaptive spatial and temporal locality-aware cache and network-on-chip (NoC) codesign (STLAC) is proposed, which dynamically partitions the last level cache (LLC) as data prefetch buffer or victim cache for locality prediction and exploits a hybrid burst-support NoC for fast data prefetch. The data prefetch buffer speculates the data blocks in subsequent addresses to exploit the spatial locality, while the victim cache collects the evicted data blocks from the upper memory hierarchy to exploit the temporal locality. By combining the proposed adaptive cache partition with the hybrid burst-support NoC, the off-chip misses and on-chip network usage are greatly reduced. Experimental results demonstrate that the proposed STLAC reduces up to 43% off-chip misses and improves 15% performance on average compared with the traditional shared LLC design.
Year
DOI
Venue
2016
10.1109/ASPDAC.2016.7427986
2016 21st Asia and South Pacific Design Automation Conference (ASP-DAC)
Keywords
Field
DocType
spatial and temporal locality-aware cache,STLAC,network-on-chip codesign,NoC codesign,tiled many-core systems,cache designs,memory wall problem,last level cache,LLC design,off-chip misses,on-chip network usage,adaptive cache partition,memory hierarchy,spatial locality,hybrid burst-support NoC,victim cache,data prefetch buffer
Computer architecture,Cache-oblivious algorithm,Cache invalidation,Locality of reference,Cache pollution,Cache,Computer science,Parallel computing,Cache algorithms,Real-time computing,Cache coloring,Instruction prefetch
Conference
ISSN
Citations 
PageRank 
2153-6961
1
0.35
References 
Authors
12
2
Name
Order
Citations
PageRank
Mingyu Wang113524.90
Zhaolin Li2289.06