Impact of data serialization over TSVs on routing congestion in 3D-stacked multi-core processors

3D integration can alleviate routing congestion, reducing the wirelength and improving performances. Nevertheless, each TSV still occupies non-negligible silicon area: as the number of TSV increases, their effect on the chip routing is detrimental. The reduction in the number of 3D vias obtained with the adoption of serial vertical connections can relieve the routing congestion of the 3D system by reducing the average wirelength. In this paper we explore the impact of the serial approach on the chip routing of a 3D multi processor platform to quantify the achievable wirelength reduction for a range of TSV technologies. The comparison between the serial and the parallel multi-processor configurations shows up to 12.4% wirelength improvement for the serial solution, with serious consequences on routing delay.

[1]  Jason Cong,et al.  A multilevel analytical placement for 3D ICs , 2009, 2009 Asia and South Pacific Design Automation Conference.

[2]  Yusuf Leblebici,et al.  3D-MMC: A modular 3D multi-core architecture with efficient resource pooling , 2013, 2013 Design, Automation & Test in Europe Conference & Exhibition (DATE).

[3]  Taewhan Kim,et al.  Algorithms for TSV resource sharing and optimization in designing 3D stacked ICs , 2014, Integr..

[4]  Kyoungrok Cho,et al.  Design of a wave-pipelined serializer-deserializer with an asynchronous protocol for high speed interfaces , 2012, 2012 4th Asia Symposium on Quality Electronic Design (ASQED).

[5]  Yusuf Leblebici,et al.  3D serial TSV link for low-power chip-to-chip communication , 2014, 2014 IEEE International Conference on IC Design & Technology.

[6]  Luca Benini,et al.  P2012: Building an ecosystem for a scalable, modular and high-efficiency embedded computing accelerator , 2012, 2012 Design, Automation & Test in Europe Conference & Exhibition (DATE).

[7]  Yusuf Leblebici,et al.  Design and Testing Strategies for Modular 3-D-Multiprocessor Systems Using Die-Level Through Silicon Via Technology , 2012, IEEE Journal on Emerging and Selected Topics in Circuits and Systems.

[8]  K. Yoshida,et al.  2.8 Gb/s 176 mW byte-interleaved and 3.0 Gb/s 118 mW bit-interleaved 8:1 multiplexers , 1996, 1996 IEEE International Solid-State Circuits Conference. Digest of TEchnical Papers, ISSCC.

[9]  Sung Kyu Lim,et al.  Through-silicon-via aware interconnect prediction and optimization for 3D stacked ICs , 2009, SLIP '09.

[10]  Uri C. Weiser,et al.  Interconnect-power dissipation in a microprocessor , 2004, SLIP '04.

[11]  Yong Liu,et al.  A compact low-power 3D I/O in 45nm CMOS , 2012, 2012 IEEE International Solid-State Circuits Conference.

[12]  Luca Benini,et al.  Design Issues and Considerations for Low-Cost 3-D TSV IC Technology , 2010, IEEE Journal of Solid-State Circuits.

[13]  Ting-Chi Wang,et al.  Through-Silicon Via Planning in 3-D Floorplanning , 2011, IEEE Transactions on Very Large Scale Integration (VLSI) Systems.

[14]  Jong-Wha Chong,et al.  Effective Estimation Method of Routing Congestion at Floorplan Stage for 3D ICs , 2011 .

[15]  Ran Ginosar,et al.  Asynchronous Current Mode Serial Communication , 2010, IEEE Transactions on Very Large Scale Integration (VLSI) Systems.