Exploring Adaptive Cache for Reconfigurable VLIW Processor

In this paper, we focus on a very long instruction word (VLIW) processor design that “shares” its cache blocks when switching to different performance modes to alleviate the aforementioned cold starts. The switching trigger cache resizing operations and improper use can lead to...

Full description

Bibliographic Details
Main Authors: Sensen Hu, Jing Haung
Format: Article
Language:English
Published: IEEE 2019-01-01
Series:IEEE Access
Subjects:
Online Access:https://ieeexplore.ieee.org/document/8725889/
id doaj-f10f76594d1c42d1ac8dff176aa301e4
record_format Article
spelling doaj-f10f76594d1c42d1ac8dff176aa301e42021-03-30T00:06:53ZengIEEEIEEE Access2169-35362019-01-017726347264610.1109/ACCESS.2019.29195898725889Exploring Adaptive Cache for Reconfigurable VLIW ProcessorSensen Hu0https://orcid.org/0000-0002-0238-7363Jing Haung1National Research Base of Intelligent Manufacturing Service, Chongqing Technology and Business University, Chongqing, ChinaSchool of Foreign Studies, Yangtze University, Jingzhou, ChinaIn this paper, we focus on a very long instruction word (VLIW) processor design that “shares” its cache blocks when switching to different performance modes to alleviate the aforementioned cold starts. The switching trigger cache resizing operations and improper use can lead to less efficient cache performance. We clearly note here that our investigation pertains the local temporal effects of the cache resizing and how we counteract the negative impact of cache misses in such resizing instances. We propose a novel reconfigurable d-cache framework that can dynamically adapt its least recently used (LRU) replacement policy without much hardware overhead. We demonstrate that using our adaptive d-cache, it ensures a smooth cache performance from one cache size to the other. This approach is orthogonal to future research in cache resizing for such architectures that take into account energy consumption and performance of the overall application. Our results show that when compared with a straightforward cache resizing approach, we can achieve a reduction of 10%-63% of cache misses during the switching period. Furthermore, we can also increase the average cache hit rate from 56% to 90% in the cache part that remains live after downsizing using our approach.https://ieeexplore.ieee.org/document/8725889/Adaptive cachecache resizereconfiguration
collection DOAJ
language English
format Article
sources DOAJ
author Sensen Hu
Jing Haung
spellingShingle Sensen Hu
Jing Haung
Exploring Adaptive Cache for Reconfigurable VLIW Processor
IEEE Access
Adaptive cache
cache resize
reconfiguration
author_facet Sensen Hu
Jing Haung
author_sort Sensen Hu
title Exploring Adaptive Cache for Reconfigurable VLIW Processor
title_short Exploring Adaptive Cache for Reconfigurable VLIW Processor
title_full Exploring Adaptive Cache for Reconfigurable VLIW Processor
title_fullStr Exploring Adaptive Cache for Reconfigurable VLIW Processor
title_full_unstemmed Exploring Adaptive Cache for Reconfigurable VLIW Processor
title_sort exploring adaptive cache for reconfigurable vliw processor
publisher IEEE
series IEEE Access
issn 2169-3536
publishDate 2019-01-01
description In this paper, we focus on a very long instruction word (VLIW) processor design that “shares” its cache blocks when switching to different performance modes to alleviate the aforementioned cold starts. The switching trigger cache resizing operations and improper use can lead to less efficient cache performance. We clearly note here that our investigation pertains the local temporal effects of the cache resizing and how we counteract the negative impact of cache misses in such resizing instances. We propose a novel reconfigurable d-cache framework that can dynamically adapt its least recently used (LRU) replacement policy without much hardware overhead. We demonstrate that using our adaptive d-cache, it ensures a smooth cache performance from one cache size to the other. This approach is orthogonal to future research in cache resizing for such architectures that take into account energy consumption and performance of the overall application. Our results show that when compared with a straightforward cache resizing approach, we can achieve a reduction of 10%-63% of cache misses during the switching period. Furthermore, we can also increase the average cache hit rate from 56% to 90% in the cache part that remains live after downsizing using our approach.
topic Adaptive cache
cache resize
reconfiguration
url https://ieeexplore.ieee.org/document/8725889/
work_keys_str_mv AT sensenhu exploringadaptivecacheforreconfigurablevliwprocessor
AT jinghaung exploringadaptivecacheforreconfigurablevliwprocessor
_version_ 1724188720246030336