The Organization of On-Chip Data Memory in One Coarse-Grained Reconfigurable Architecture

Yansheng WANG; Leibo LIU; Shouyi YIN; Min ZHU; Peng CAO; Jun YANG; Shaojun WEI

doi:10.1587/transfun.E96.A.2218

The Organization of On-Chip Data Memory in One Coarse-Grained Reconfigurable Architecture

Yansheng WANG, Leibo LIU, Shouyi YIN, Min ZHU, Peng CAO, Jun YANG, Shaojun WEI

Full Text Views

0

Share
Cite this

Summary :

RCP (Reconfigurable Computing Processor) is intended to fill the gap between ASIC and GPP (General Purpose processor), which achieves much higher energy efficiency than GPP, while is much more flexible than ASIC. In this paper, one organization of on-chip data memory called LIBODM (LIfetime Based On-chip Data Memory) is proposed to reduce the reference delay for data and on-chip data memory size in RCP. In the LIBODM, the allocation of data is based on the data dependency. The data with low data dependency are stored off-chip to save the storage costs, while the data with high data dependency are stored on-chip to reduce the reference delay. Besides, in the LIBODM, the on-chip data are classified into two types, and the classification is based on the lifetime of data. For short lifetime data, they are preferred to be stored into FIFO to increase the reuse ratio of memory space naturally. For long lifetime data, they are preferred to be stored into RAM for several time references. The LIBODM has been testified in one CGRA (Coarse Grained Reconfigurable Architecture) called RPU (Reconfigurable Processing Unit), and two RPUs has been integrated in a RCP-REMUS_HP (High Performance version of Reconfigurable MUlti-media System) focused on video decoding. Thanks to the LIBODM, although the size of on-chip data memory in REMUS_HP is small, a high performance can still be achieved. Compared with XPP and ADRES, in REMUS_HP, the on-chip data memory size at same performance level is only 23.9% and 14.8%. REMUS_HP is implemented on a 48.9mm² silicon with TSMC 65nm technology. Simulation shows that 1920*1088 @30fps can be achieved for H.264 high-profile decoding when exploiting a 200MHz working frequency. Compared with the high performance version of XPP, the performance is 150% boosted, while the energy efficiency is 17.59x boosted.

Publication: IEICE TRANSACTIONS on Fundamentals Vol.E96-A No.11 pp.2218-2229

Publication Date: 2013/11/01

Publicized

Online ISSN: 1745-1337

DOI: 10.1587/transfun.E96.A.2218

Type of Manuscript: PAPER

Category: VLSI Design Technology and CAD

Authors

Yansheng WANG
  Tsinghua University
Leibo LIU
  Tsinghua University
Shouyi YIN
  Tsinghua University
Min ZHU
  Tsinghua University
Peng CAO
  Southeast Univ.
Jun YANG
  Southeast Univ.
Shaojun WEI
  Tsinghua University

Keyword

dynamic reconfigurable, on-chip data memory, fast reference, low storage costs

Cite this

Copy

Yansheng WANG, Leibo LIU, Shouyi YIN, Min ZHU, Peng CAO, Jun YANG, Shaojun WEI, "The Organization of On-Chip Data Memory in One Coarse-Grained Reconfigurable Architecture" in IEICE TRANSACTIONS on Fundamentals, vol. E96-A, no. 11, pp. 2218-2229, November 2013, doi: 10.1587/transfun.E96.A.2218.
Abstract: RCP (Reconfigurable Computing Processor) is intended to fill the gap between ASIC and GPP (General Purpose processor), which achieves much higher energy efficiency than GPP, while is much more flexible than ASIC. In this paper, one organization of on-chip data memory called LIBODM (LIfetime Based On-chip Data Memory) is proposed to reduce the reference delay for data and on-chip data memory size in RCP. In the LIBODM, the allocation of data is based on the data dependency. The data with low data dependency are stored off-chip to save the storage costs, while the data with high data dependency are stored on-chip to reduce the reference delay. Besides, in the LIBODM, the on-chip data are classified into two types, and the classification is based on the lifetime of data. For short lifetime data, they are preferred to be stored into FIFO to increase the reuse ratio of memory space naturally. For long lifetime data, they are preferred to be stored into RAM for several time references. The LIBODM has been testified in one CGRA (Coarse Grained Reconfigurable Architecture) called RPU (Reconfigurable Processing Unit), and two RPUs has been integrated in a RCP-REMUS_HP (High Performance version of Reconfigurable MUlti-media System) focused on video decoding. Thanks to the LIBODM, although the size of on-chip data memory in REMUS_HP is small, a high performance can still be achieved. Compared with XPP and ADRES, in REMUS_HP, the on-chip data memory size at same performance level is only 23.9% and 14.8%. REMUS_HP is implemented on a 48.9mm² silicon with TSMC 65nm technology. Simulation shows that 1920*1088 @30fps can be achieved for H.264 high-profile decoding when exploiting a 200MHz working frequency. Compared with the high performance version of XPP, the performance is 150% boosted, while the energy efficiency is 17.59x boosted.
URL: https://globals.ieice.org/en_transactions/fundamentals/10.1587/transfun.E96.A.2218/_p

Copy

@ARTICLE{e96-a_11_2218,
author={Yansheng WANG, Leibo LIU, Shouyi YIN, Min ZHU, Peng CAO, Jun YANG, Shaojun WEI, },
journal={IEICE TRANSACTIONS on Fundamentals},
title={The Organization of On-Chip Data Memory in One Coarse-Grained Reconfigurable Architecture},
year={2013},
volume={E96-A},
number={11},
pages={2218-2229},
abstract={RCP (Reconfigurable Computing Processor) is intended to fill the gap between ASIC and GPP (General Purpose processor), which achieves much higher energy efficiency than GPP, while is much more flexible than ASIC. In this paper, one organization of on-chip data memory called LIBODM (LIfetime Based On-chip Data Memory) is proposed to reduce the reference delay for data and on-chip data memory size in RCP. In the LIBODM, the allocation of data is based on the data dependency. The data with low data dependency are stored off-chip to save the storage costs, while the data with high data dependency are stored on-chip to reduce the reference delay. Besides, in the LIBODM, the on-chip data are classified into two types, and the classification is based on the lifetime of data. For short lifetime data, they are preferred to be stored into FIFO to increase the reuse ratio of memory space naturally. For long lifetime data, they are preferred to be stored into RAM for several time references. The LIBODM has been testified in one CGRA (Coarse Grained Reconfigurable Architecture) called RPU (Reconfigurable Processing Unit), and two RPUs has been integrated in a RCP-REMUS_HP (High Performance version of Reconfigurable MUlti-media System) focused on video decoding. Thanks to the LIBODM, although the size of on-chip data memory in REMUS_HP is small, a high performance can still be achieved. Compared with XPP and ADRES, in REMUS_HP, the on-chip data memory size at same performance level is only 23.9% and 14.8%. REMUS_HP is implemented on a 48.9mm² silicon with TSMC 65nm technology. Simulation shows that 1920*1088 @30fps can be achieved for H.264 high-profile decoding when exploiting a 200MHz working frequency. Compared with the high performance version of XPP, the performance is 150% boosted, while the energy efficiency is 17.59x boosted.},
keywords={},
doi={10.1587/transfun.E96.A.2218},
ISSN={1745-1337},
month={November},}

Copy

TY - JOUR
TI - The Organization of On-Chip Data Memory in One Coarse-Grained Reconfigurable Architecture
T2 - IEICE TRANSACTIONS on Fundamentals
SP - 2218
EP - 2229
AU - Yansheng WANG
AU - Leibo LIU
AU - Shouyi YIN
AU - Min ZHU
AU - Peng CAO
AU - Jun YANG
AU - Shaojun WEI
PY - 2013
DO - 10.1587/transfun.E96.A.2218
JO - IEICE TRANSACTIONS on Fundamentals
SN - 1745-1337
VL - E96-A
IS - 11
JA - IEICE TRANSACTIONS on Fundamentals
Y1 - November 2013
AB - RCP (Reconfigurable Computing Processor) is intended to fill the gap between ASIC and GPP (General Purpose processor), which achieves much higher energy efficiency than GPP, while is much more flexible than ASIC. In this paper, one organization of on-chip data memory called LIBODM (LIfetime Based On-chip Data Memory) is proposed to reduce the reference delay for data and on-chip data memory size in RCP. In the LIBODM, the allocation of data is based on the data dependency. The data with low data dependency are stored off-chip to save the storage costs, while the data with high data dependency are stored on-chip to reduce the reference delay. Besides, in the LIBODM, the on-chip data are classified into two types, and the classification is based on the lifetime of data. For short lifetime data, they are preferred to be stored into FIFO to increase the reuse ratio of memory space naturally. For long lifetime data, they are preferred to be stored into RAM for several time references. The LIBODM has been testified in one CGRA (Coarse Grained Reconfigurable Architecture) called RPU (Reconfigurable Processing Unit), and two RPUs has been integrated in a RCP-REMUS_HP (High Performance version of Reconfigurable MUlti-media System) focused on video decoding. Thanks to the LIBODM, although the size of on-chip data memory in REMUS_HP is small, a high performance can still be achieved. Compared with XPP and ADRES, in REMUS_HP, the on-chip data memory size at same performance level is only 23.9% and 14.8%. REMUS_HP is implemented on a 48.9mm² silicon with TSMC 65nm technology. Simulation shows that 1920*1088 @30fps can be achieved for H.264 high-profile decoding when exploiting a 200MHz working frequency. Compared with the high performance version of XPP, the performance is 150% boosted, while the energy efficiency is 17.59x boosted.
ER -