Dynamic directory table with victim cache: on-demand allocation of directory entries for active shared cache blocks

被引:0
|
作者
Han Jun Bae
Lynn Choi
机构
[1] Korea University,School of Electrical Engineering
来源
关键词
Cache coherence; Directory; Parallel processing; Simulation; Scalable computing; Multi-core architectures;
D O I
暂无
中图分类号
学科分类号
摘要
In this paper, we present a novel directory architecture that can dynamically allocate a directory entry for a cache block on demand at runtime only when the block is shared by more than a single core. Thus, we do not maintain coherence for private blocks, substantially reducing the number of directory entries. Even for shared blocks, we allocate directory entry dynamically only when the block is actively shared, further reducing the number of directory entries at runtime. For this, we propose a new directory architecture called dynamic directory table (DDT), which is a decoupled directory storage from the shared cache and dynamically maintains directory entries only for actively shared blocks. Also, we add a small additional victim cache to its original DDT in order to reduce invalidation broadcasts caused by DDT eviction. Through our detailed simulation on PARSEC benchmarks, we show that DDT can outperform the expensive full-map directory by a slight margin with only 16.09% of directory area across a variety of different workloads. This is achieved by its faster access and high hit rates in the small directory. In addition, we demonstrate that even smaller DDTs can give comparable or higher performance compared to recent directory optimization schemes such as SPACE and DGD with considerably less area.
引用
收藏
页码:425 / 446
页数:21
相关论文
共 10 条
  • [1] Dynamic directory table with victim cache: on-demand allocation of directory entries for active shared cache blocks
    Bae, Han Jun
    Choi, Lynn
    [J]. JOURNAL OF SUPERCOMPUTING, 2019, 75 (01): : 425 - 446
  • [2] A Directory Cache with Dynamic Private-Shared Partitioning
    Valls, Joan J.
    Gomez, Maria E.
    Ros, Alberto
    Sahuquillo, Julio
    [J]. PROCEEDINGS OF 2016 IEEE 23RD INTERNATIONAL CONFERENCE ON HIGH PERFORMANCE COMPUTING (HIPC), 2016, : 382 - 391
  • [3] Zero Directory Eviction Victim: Unbounded Coherence Directory and Core Cache Isolation
    Chaudhuri, Mainak
    [J]. 2021 27TH IEEE INTERNATIONAL SYMPOSIUM ON HIGH-PERFORMANCE COMPUTER ARCHITECTURE (HPCA 2021), 2021, : 277 - 290
  • [4] Selective cache ways: On-demand cache resource allocation
    Albonesi, DH
    [J]. 32ND ANNUAL INTERNATIONAL SYMPOSIUM ON MICROARCHITECTURE, (MICRO-32), PROCEEDINGS, 1999, : 248 - 259
  • [5] Directory Lookaside Table: Enabling Scalable, Low-Conflict, Many-Core Cache Coherence Directory
    Shi, Xudong
    Su, Feiqi
    Peir, Jih-Kwon
    [J]. 2014 20TH IEEE INTERNATIONAL CONFERENCE ON PARALLEL AND DISTRIBUTED SYSTEMS (ICPADS), 2014, : 111 - 118
  • [6] A Novel Directory Based Hybrid Cache Coherence Protocol for Shared Memory Multiprocessors
    Asaduzzaman, Abu
    Chidella, Kishore K.
    [J]. 2016 IEEE INTERNATIONAL SYMPOSIUM ON PHASED ARRAY SYSTEMS AND TECHNOLOGY (PAST), 2016,
  • [7] Dynamic Last-Level Cache Allocation to Reduce Area and Power Overhead in Directory Coherence Protocols
    Lodde, Mario
    Flich, Jose
    Acacio, Manuel E.
    [J]. EURO-PAR 2012 PARALLEL PROCESSING, 2012, 7484 : 206 - 218
  • [8] On-demand data co-allocation with user-level cache for grids
    Chen, Po-Cheng
    Chang, Jyh-Biau
    Su, Yen-Liang
    Shieh, Ce-Kuen
    [J]. CONCURRENCY AND COMPUTATION-PRACTICE & EXPERIENCE, 2010, 22 (18): : 2488 - 2513
  • [9] Optimal Bandwidth Allocation for Multicast-Cache-Aided on-Demand Streaming in Wireless Networks
    Amidzadeh, Mohsen
    Tirkkonen, Olav
    Caire, Giuseppe
    [J]. 2022 IEEE GLOBAL COMMUNICATIONS CONFERENCE (GLOBECOM 2022), 2022, : 1936 - 1941
  • [10] Boosting Performance of Directory-based Cache Coherence Protocols with Coherence Bypass at Subpage Granularity and A Novel On-chip Page Table
    Soltaniyeh, Mohammadreza
    Kadayif, Ismail
    Ozturk, Ozcan
    [J]. PROCEEDINGS OF THE ACM INTERNATIONAL CONFERENCE ON COMPUTING FRONTIERS (CF'16), 2016, : 180 - 187