Low-Cost Strategy for Bus Propagation Delay Reduction

We propose a strategy to reduce the propagation delay of microprocessors’ digital bus lines at very low costs in terms of area overhead, power consumption and power-delay product. Likewise some solutions adopted in industry nowadays, our strategy inserts in the bus lines repeaters implemented as a chain of inverters with increasing size. In this paper, we derive new expressions to determine the optimum number of inverters to be used within each repeater, and the optimum number of repeaters to insert in the bus lines. Our derived expressions yield to bus implementations with significant lower cost in terms of area overhead and power consumption than alternative solutions in literature. Considering a 32 nm technology as a significant example, we show that, compared to the traditional solution that inserts repeaters implemented as a single inverter, our strategy enables reductions up to 84% in terms of area overhead, up to 65% in power consumption, and up to 66% in power-delay product. Compared to three recent alternative solutions in the literature, our strategy enables reductions up to 88% in terms of area overhead, up to 48% in power consumption, and up to 43% in power-delay product. Therefore, our approach is particularly suitable to the growing market of mobile applications that require low cost in terms of power consumption and chip area.

[1]  Cecilia Metra,et al.  Function-Inherent Code Checking: A New Low Cost On-Line Testing Approach for High Performance Microprocessor Control Logic , 2008, 2008 13th European Test Symposium.

[2]  Andrew B. Kahng,et al.  Interconnect optimization strategies for high-performance VLSI designs , 1999, Proceedings Twelfth International Conference on VLSI Design. (Cat. No.PR00013).

[3]  Magdy A. Bayoumi,et al.  Reducing Interconnect Delay Uncertainty via Hybrid Polarity Repeater Insertion , 2008, IEEE Transactions on Very Large Scale Integration (VLSI) Systems.

[4]  David Blaauw,et al.  Modeling and analysis of crosstalk noise in coupled RLC interconnects , 2006, IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems.

[5]  Nasser Masoumi,et al.  An analytical delay reduction strategy for buffer-inserted global interconnects in VDSM technologies , 2009, 2009 European Conference on Circuit Theory and Design.

[6]  Cecilia Metra,et al.  Exploiting ECC redundancy to minimize crosstalk impact , 2005, IEEE Design & Test of Computers.

[7]  Daniel Mossé,et al.  Optimized Management of Power and Performance for Virtualized Heterogeneous Server Clusters , 2011, 2011 11th IEEE/ACM International Symposium on Cluster, Cloud and Grid Computing.

[8]  Xuebin Wu,et al.  Efficient CODEC Designs for Crosstalk Avoidance Codes Based on Numeral Systems , 2011, IEEE Transactions on Very Large Scale Integration (VLSI) Systems.

[9]  Luca Benini,et al.  Analysis of error recovery schemes for networks on chips , 2005, IEEE Design & Test of Computers.

[10]  A. Karthikeyan,et al.  High-speed and low-power repeater for VLSI interconnects , 2017 .

[11]  A. Karthikeyan,et al.  Optimization Techniques for CNT Based VLSI Interconnects - A Review , 2017, J. Circuits Syst. Comput..

[12]  Eby G. Friedman,et al.  Optimum wire sizing of RLC interconnect with repeaters , 2004, Integr..

[13]  Yehea Ismail,et al.  Optimum positioning of interleaved repeaters in bidirectional buses , 2005, IEEE Trans. Comput. Aided Des. Integr. Circuits Syst..

[14]  Naresh R. Shanbhag,et al.  Coding for reliable on-chip buses: a class of fundamental bounds and practical codes , 2007, IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems.

[15]  Amirreza Alizadeh,et al.  On Temperature Dependency of Delay for Local,Intermediate, and Repeater Inserted Global Copper Interconnects , 2015, IEEE Transactions on Very Large Scale Integration (VLSI) Systems.

[16]  S. Muddu,et al.  Repeater and interconnect strategies for high-performance physical designs , 1998, Proceedings. XI Brazilian Symposium on Integrated Circuit Design (Cat. No.98EX216).