TY - GEN
T1 - MuCCRA chips
T2 - 2007 IEEE Asian Solid-State Circuits Conference, A-SSCC
AU - Amano, Hideharu
AU - Hasegawa, Yohei
AU - Tsutsumi, Satoshi
AU - Nakamura, Takuro
AU - Nishimura, Takashi
AU - Tanbunheng, Vasutan
AU - Parimala, Aepu
AU - Sano, Toru
AU - Kato, Masaru
PY - 2007/12/1
Y1 - 2007/12/1
N2 - Coarse grained dynamically reconfigurable processor arrays (DRPAs) have been received an attention as a flexible and efficient off-loading engine in System-On-Chips (SoCs). Evaluation results in recent researches revealed that the parameters of optimal processor array structure: granularity, functions, array size, context size and interconnection flexibility, are completely different for each application. That is, DRPAs should be configurable for target SoCs and applications. MuCCRA is a project for developing a DRPA generator which can generate RTL model, testing environment and programming environment for various types of DRPAs just by selecting the specific parameters. Here, two prototype chips MuCCRA-1 and MuCCRA-2 developed in the project are introduced and evaluated. MuCCRA-1 was implemented with Rohm's 0.18um CMOS process mainly for multi-media applications, while MuCCRA-2 with ASPLA's 90nm CMOS process was designed focusing on area optimization used as a cost-effective IP in multi-core SoCs.
AB - Coarse grained dynamically reconfigurable processor arrays (DRPAs) have been received an attention as a flexible and efficient off-loading engine in System-On-Chips (SoCs). Evaluation results in recent researches revealed that the parameters of optimal processor array structure: granularity, functions, array size, context size and interconnection flexibility, are completely different for each application. That is, DRPAs should be configurable for target SoCs and applications. MuCCRA is a project for developing a DRPA generator which can generate RTL model, testing environment and programming environment for various types of DRPAs just by selecting the specific parameters. Here, two prototype chips MuCCRA-1 and MuCCRA-2 developed in the project are introduced and evaluated. MuCCRA-1 was implemented with Rohm's 0.18um CMOS process mainly for multi-media applications, while MuCCRA-2 with ASPLA's 90nm CMOS process was designed focusing on area optimization used as a cost-effective IP in multi-core SoCs.
UR - http://www.scopus.com/inward/record.url?scp=51349116937&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=51349116937&partnerID=8YFLogxK
U2 - 10.1109/ASSCC.2007.4425711
DO - 10.1109/ASSCC.2007.4425711
M3 - Conference contribution
AN - SCOPUS:51349116937
SN - 1424413605
SN - 9781424413607
T3 - 2007 IEEE Asian Solid-State Circuits Conference, A-SSCC
SP - 384
EP - 387
BT - 2007 IEEE Asian Solid-State Circuits Conference, A-SSCC
Y2 - 12 November 2007 through 14 November 2007
ER -