Variation-Tolerant and Low-Power Source-Synchronous Multicycle On-Chip Interconnect Scheme

被引:7
作者
Ghoneima, Maged [1 ]
Ismail, Yehea [2 ]
Khellah, Muhammad [3 ]
De, Vivek [3 ]
机构
[1] NVIDA Corp, VLSI Design Gronp, Santa Clara, CA 95050 USA
[2] Northwestern Univ, Elect Enginering & Comp Sci Dept EECS, Evanston, IL 60208 USA
[3] Intel Corp, Circuit Res Labs, Hillsboro, OR 97124 USA
关键词
D O I
10.1155/2007/95402
中图分类号
TP3 [计算技术、计算机技术];
学科分类号
0812 ;
摘要
A variation-tolerant low-power source-synchronous multicycle (SSMC) interconnect scheme is proposed. This scheme is scalable and suitable for transferring data across different clock domains such as those in " many-core" SoCs and in 3D-ICs. SSMC replaces intermediate flip-flops by a source-synchronous synchronization scheme. Removing the intermediate flip-flops in the SSMC scheme enables better averaging of delay variations across the whole interconnect, which reduces bit-rate degradation due to within-die WID process variations. Monte Carlo circuit simulations show that SSMC eliminates 90% of the variation-induced performance degradation in a 6-cycle 9 mm-long 16-bit conventional bus. The proposedmulticycle bus scheme also leads to significant energy savings due to eliminating the power-hungry flip-flops and efficiently designing the source synchronization overhead. Moreover, eliminating intermediate flip-flops avoids the timing overhead of the setup time, the flip-flop delay, and the single-cycle clock jitter. This delay slack can then be translated into further energy savings by downsizing the repeaters. The significant delay jitter due to capacitive coupling has been addressed and solutions are put forward to alleviate it. Circuit simulations in a 65-nm process environment indicate that energy savings up to 20% are achievable for a 6-cycle 9mm long 16-bit bus. Copyright (C) 2007 Maged Ghoneima et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
引用
收藏
页数:12
相关论文
共 23 条
[1]  
[Anonymous], 2013, 114912013 IEEE, P1
[2]   OPTIMAL INTERCONNECTION CIRCUITS FOR VLSI [J].
BAKOGLU, HB ;
MEINDL, JD .
IEEE TRANSACTIONS ON ELECTRON DEVICES, 1985, 32 (05) :903-909
[3]   A hybrid current/voltage mode on-chip signaling scheme with adaptive bandwidth capability [J].
Bashirullah, R ;
Liu, WT ;
Cavin, R ;
Edwards, D .
IEEE TRANSACTIONS ON VERY LARGE SCALE INTEGRATION (VLSI) SYSTEMS, 2004, 12 (08) :876-880
[4]   3D processing technology and its impact on iA32 microprocessors [J].
Black, B ;
Nelson, DW ;
Webb, C ;
Samra, N .
IEEE INTERNATIONAL CONFERENCE ON COMPUTER DESIGN: VLSI IN COMPUTERS & PROCESSORS, PROCEEDINGS, 2004, :316-318
[5]  
Borkar S., 2015, CISC VIS NETW IND GL
[6]   Theory of latency-insensitive design [J].
Carloni, LP ;
McMillan, KL ;
Sangiovanni-Vincentelli, AL .
IEEE TRANSACTIONS ON COMPUTER-AIDED DESIGN OF INTEGRATED CIRCUITS AND SYSTEMS, 2001, 20 (09) :1059-1076
[7]   Concurrent flip-flop and repeater insertion for high performance integrated circuits [J].
Cocchini, P .
IEEE/ACM INTERNATIONAL CONFERENCE ON CAD-02, DIGEST OF TECHNICAL PAPERS, 2002, :268-273
[8]  
Dally W., 1998, DIGITAL SYSTEMS ENG
[9]   Three-dimensional integrated circuits: Performance, design methodology, and CAD tools [J].
Das, S ;
Chandrakasan, A ;
Reif, R .
ISVLSI 2003: IEEE COMPUTER SOCIETY ANNUAL SYMPOSIUM ON VLSI, PROCEEDINGS: NEW TRENDS AND TECHNOLOGIES FOR VLSI SYSTEMS DESIGN, 2003, :13-18
[10]   Timing closure through a globally synchronous, timing partitioned design methodology [J].
Edman, A ;
Svensson, C .
41ST DESIGN AUTOMATION CONFERENCE, PROCEEDINGS 2004, 2004, :71-74