MRCA: Multi-grained Reconfigurable Cryptographic Accelerator for Diverse Security Requirements

被引:0
|
作者
Pham Hoai Luan [1 ]
Hai Hau Nguyen [2 ]
Vu Trung Duong Le [1 ]
Thi Diem Tran [2 ]
Tuan Hai Vu [1 ]
Thi Hong Tran [3 ]
Yasuhiko Nakashima [1 ]
机构
[1] Nara Inst Sci & Technol, Nara, Japan
[2] Univ Informat Technol VNUHCM, Ho Chi Minh City, Vietnam
[3] Osaka Metropolitan Univ, Osaka 5588585, Japan
关键词
CGRA; Cryptography; FPGA; Accelerator; PROCESSOR;
D O I
10.1109/COOLCHIPS61292.2024.10531185
中图分类号
TP3 [计算技术、计算机技术];
学科分类号
0812 ;
摘要
In today's digital security, the robust 64-bit cryptographic algorithms are key to fulfilling strict security standards. However, most existing cryptographic architectures primarily support 8/32-bit algorithms and often exhibit poor performance. Therefore, this paper proposes a multi-grained reconfigurable cryptographic accelerator (MRCA) to efficiently support a variety of 8/32/64-bit algorithms with high performance and energy efficiency. To achieve this, we propose three innovative ideas. Firstly, a dual processing element array (D-PEA) is proposed to perform either two parallel 32-bit algorithms or concatenate them for 64-bit algorithms. Secondly, we develop a synchronous row connection and processing element architecture with a double buffer lane to reduce context memory and enable flexible coordination and computation of data in both 32-bit and 64-bit algorithms. Thirdly, we propose a concatenable 32-bit crypto ALU within each processing element, which can not only execute multiple 32-bit cryptographic operations in a single cycle but also connect to support multiple 64-bit operations. Our MRCA has been successfully implemented and verified across various 8/32/64-bit cryptographic algorithms on the TySOM-3A FPGA platform. Our ASIC experiment on a 45nm CMOS technology shows that the MRCA achieves a remarkable throughput from 0.43 to 15.48 Gbps and energy efficiency ranging from 0.4 to 14.46 Gbps/W, covering a wide range of 8-bit, 32-bit, and 64-bit cryptographic algorithm computations. Compared to previous studies, the MRCA outperforms by 3.5-12 times in throughput and 3.7-20 times in energy efficiency for hashing computations, also offering enhanced flexibility in supporting a wider range of 64-bit algorithms.
引用
收藏
页数:6
相关论文
共 13 条
  • [1] A Multi-Grained Reconfigurable Accelerator for Approximate Computing
    Kan, Yirong
    Wu, Man
    Zhang, Renyuan
    Nakashima, Yasuhiko
    2020 IEEE COMPUTER SOCIETY ANNUAL SYMPOSIUM ON VLSI (ISVLSI 2020), 2020, : 90 - 95
  • [2] MuGRA: A Scalable Multi-Grained Reconfigurable Accelerator Powered by Elastic Neural Network
    Kan, Yirong
    Wu, Man
    Zhang, Renyuan
    Nakashima, Yasuhiko
    IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS I-REGULAR PAPERS, 2022, 69 (01) : 258 - 271
  • [3] LiCryptor: High-Speed and Compact Multi-Grained Reconfigurable Accelerator for Lightweight Cryptography
    Pham, Hoai Luan
    Le, Vu Trung Duong
    Tran, Van Duy
    Vu, Tuan Hai
    Nakashima, Yasuhiko
    IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS I-REGULAR PAPERS, 2024, 71 (10) : 4624 - 4637
  • [4] Exploiting multi-grained parallelism in reconfigurable SBC architectures
    Zambreno, J
    Honbo, D
    Choudhary, A
    FCCM 2005: 13TH ANNUAL IEEE SYMPOSIUM ON FIELD-PROGRAMMABLE CUSTOM COMPUTING MACHINES, PROCEEDINGS, 2005, : 333 - 334
  • [5] MSCA: A Multi-grained Sparse Convolution Accelerator for DNN Training
    Mao, Yingchang
    Liu, Qiang
    Cheung, Ray C. C.
    2024 IEEE 35TH INTERNATIONAL CONFERENCE ON APPLICATION-SPECIFIC SYSTEMS, ARCHITECTURES AND PROCESSORS, ASAP 2024, 2024, : 34 - 35
  • [6] Multi-grained reconfigurable datapath structures for online-adaptive reconfigurable hardware architectures
    Thomas, A
    Becker, J
    IEEE COMPUTER SOCIETY ANNUAL SYMPOSIUM ON VLSI, PROCEEDINGS: NEW FRONTIERS IN VLSI DESIGN, 2005, : 118 - 123
  • [7] HoneyComb: A Multi-grained Dynamically Reconfigurable Runtime Adaptive Hardware Architecture
    Thomas, Alexander
    Rueckauer, Michael
    Becker, Juergen
    2011 IEEE INTERNATIONAL SOC CONFERENCE (SOCC), 2011, : 335 - 340
  • [8] An Adaptive Cryptographic Accelerator for Network Storage Security on Dynamically Reconfigurable Platform
    Tang, Li
    Liu, Jing-Ning
    Feng, Dan
    Tong, Wei
    EIGHTH INTERNATIONAL SYMPOSIUM ON OPTICAL STORAGE AND 2008 INTERNATIONAL WORKSHOP ON INFORMATION DATA STORAGE, 2009, 7125
  • [9] Multi-Grained Trace Collection, Analysis, and Management of Diverse Container Images
    Huang, Zhuo
    Zhang, Qi
    Fan, Hao
    Wu, Song
    Yu, Chen
    Jin, Hai
    Deng, Jun
    Gu, Jing
    Tang, Zhimin
    IEEE TRANSACTIONS ON COMPUTERS, 2024, 73 (07) : 1698 - 1710
  • [10] An Elastic Neural Network Toward Multi-Grained Re-configurable Accelerator
    Wu, Man
    Chen, Yan
    Ran, Yirong
    Nomura, Takeshi
    Zhang, Renyuan
    Nakashima, Yasuhiko
    2020 18TH IEEE INTERNATIONAL NEW CIRCUITS AND SYSTEMS CONFERENCE (NEWCAS'20), 2020, : 218 - 221