Performance Comparisons Between Carbon Nanotubes, Optical, and Cu for Future High-Performance On-Chip Interconnect Applications

Optical interconnects and carbon nanotubes (CNTs) present promising options for replacing the existing Cu-based global/semiglobal (optics and CNT) and local (CNT) wires. We quantify the performance of these novel interconnects and compare it with Cu/low-kappa wires for future high-performance integrated circuits. We find that for a local wire, a CNT bundle exhibits a smaller latency than Cu for a given geometry. In addition, by leveraging the superior electromigration properties of CNT and optimizing its geometry, the latency advantage can be further amplified. For semiglobal and global wires, we compare both optical and CNT options with Cu in terms of latency, energy efficiency/power dissipation, and bandwidth density. The above trends are studied with technology node. In addition, for a future technology node, we compare the relationship between bandwidth density, power density, and latency, thus alluding to the latency and power penalty to achieve a given bandwidth density. Optical wires have the lowest latency and the highest possible bandwidth density using wavelength division multiplexing, whereas a CNT bundle has a lower latency than Cu. The power density comparison is highly switching activity (SA) dependent, with high SA favoring optics. At low SA, optics is only power efficient compared to CNT for a bandwidth density beyond a critical value. Finally, we also quantify the impact of improvement in optical and CNT technology on the above comparisons. A small monolithically integrated detector and modulator capacitance for optical interconnects (~10 fF) yields a superior power density and latency even at relatively lower SA (~20%) but at high bandwidth density. At lower bandwidth density and SA lower than 20%, an improvement in mean free path and packing density of CNT can render it most energy efficient.     

Compact physical models for multiwall carbon-nanotube interconnects

Compact physical models are derived for conductivity of multiwall carbon-nanotube (MWCN) interconnects. It is proven that for MWCNs shorter than the critical length (typically around 7 /spl mu/m), the conductivity decreases as diameter increases, whereas for MWCNs longer than the critical length, increasing the diameter results in higher conductivities. For long lengths (hundreds of micrometers), MWCNs can potentially have conductivities several times larger than that of copper or even single-wall carbon nanotube (SWCN) bundles. For short lengths (<10 /spl mu/m), however, SWCN bundles offer more than two times higher conductivities compared to MWCNs.     

Global interconnect width and spacing optimization for latency, bandwidth and power dissipation

This paper addresses a novel methodology optimizing global interconnect width and spacing for International Technology Roadmap for Semiconductors technology nodes. Global interconnects with and without buffer insertion are considered. The effects of the width and spacing of global interconnects on performance, such as delay, bandwidth, total repeater area and energy dissipation, are analyzed. The product of delay and bandwidth is used as the figure of merit for simultaneous short latency and large bandwidth and the proposed methodology can optimize global interconnects for the maximal figure of merit. It is demonstrated that buffers should not be inserted in global interconnects if interconnect length is shorter than a critical length, which is a constant for a given technology. For global interconnects with buffer insertion, the optimal width and spacing have analytical expressions and are constants for a given technology. For global interconnects without buffer insertion, the optimal width and spacing are dependent on both the technology parameters and interconnect length and can be computed numerically.     

基于单壁碳纳米管束的三维互连线设计与分析

Metallic carbon nanotubes(CNTs) have been proposed as a promising alternative to Cu interconnects in future integrated circuits(ICs) for their remarkable conductive, mechanical and thermal properties. Compact equivalent circuit models for single-walled carbon nanotube(SWCNT) bundles are described, and the performance of SWCNT bundle interconnects is evaluated and compared with traditional Cu interconnects at different interconnect levels for through-silicon-via-based three dimensional(3D) ICs. It is shown that at a local level, CNT interconnects exhibit lower signal delay and smaller optimal wire size. At intermediate and global levels, the delay improvement becomes more significant with technology scaling and increasing wire lengths. For 1 mm intermediate and 10 mm global level interconnects, the delay of SWCNT bundles is only 49.49% and 52.82% that of the Cu wires, respectively.     

Monolayer metallic nanotube interconnects: promising candidates for short local interconnects

Mono- or bi-layer metallic single-wall carbon nanotube interconnects have lateral capacitances more than four times smaller than those of copper interconnects. The resistance and time-of-flight of these monolayer nanotubes would be larger than that of copper interconnects. For short lengths, however, driver resistance is quite dominant, and latency is determined by interconnect capacitance. Monolayer nanotube interconnects are therefore promising candidates for local interconnects. The average capacitance per unit length of these nanotube interconnects can be 50% smaller than that of copper interconnects and that leads to significant saving in power dissipation.     

Compact Performance Models and Comparisons for Gigascale On-Chip Global Interconnect Technologies

The on-chip global interconnect with conventional Cu/low-k and delay-optimized repeater scheme faces great challenges in the nanometer regime owing to its severe performance degradation. This paper describes the analytical models and performance comparisons of novel interconnect technologies and circuit architectures to cope with the interconnect performance bottlenecks. Carbon nanotubes (CNTs) and optics-based interconnects exhibit promising physical properties for replacing the current Cu/low-k-based global interconnects. We quantify the performance of these novel interconnects and compare them with Cu/low-k wires for future high-performance integrated circuits. The foregoing trends are studied with technology node and bandwidth density in terms of latency and power dissipation. Optical wires have the lowest latency and power consumption, whereas a CNT bundle has a lower latency than Cu. The new circuit scheme, i.e., “capacitively driven low-swing interconnect (CDLSI),” has the potential to effect a significant energy saving and latency reduction. We present an accurate analytical optimization model for the CDLSI wire scheme. In addition, we quantify and compare the delay and energy expenditure for not only the different interconnect circuit schemes but also the various future technologies, such as Cu, CNT, and optics. We find that the CDLSI circuit scheme outperforms the conventional interconnects in latency and energy per bit for a lower bandwidth requirement, whereas these advantages degrade for higher bandwidth requirements. Finally, we explore the impact of the CNT bundle and the CDLSI on a via blockage factor. The CNT shows a significant reduction in via blockage, whereas the CDLSI does not help to alleviate it, although the CDLSI results in a reduced number of repeaters due to the differential signaling scheme.      

Design and Optimization of On-Chip Interconnects Using Wave-Pipelined Multiplexed Routing

Every new VLSI technology generation has resulted in interconnects increasingly limiting the performance, area, and power dissipation of new processors. Subsequently, it is necessary to devise efficient interconnect design techniques to reduce the impact of VLSI interconnects on overall system design. New optimizations of a wave-pipelined multiplexed (WPM) interconnect routing circuit are described in this paper. These WPM circuits can be used with current interconnect repeater circuits to further reduce interconnect delay, interconnect area, transistor area, and/or power dissipation. For example, new area constrained WPM circuit optimizations illustrate that the interconnect circuit power can be reduced by 26% or the interconnect performance can be improved by 74%. Moreover, in both these cases, because a significant number of repeaters are eliminated, the transistor area can reduce by 41% or 29%, respectively. Finally, the tolerance of WPM circuits to crosstalk noise, power supply noise, clock skew, and manufacturing variations is also presented. This study of tolerance levels defines the conditions under which the WPM circuit will function correctly, and it is shown in this paper for the first time that WPM circuits are robust enough to operate with variability that can be encountered in deep submicrometer technologies.     

Circuit Modeling and Performance Analysis of Multi-Walled Carbon Nanotube Interconnects

Metallic carbon nanotubes (CNTs) have received much attention for their unique characteristics as a possible alternative to Cu interconnects in future ICs. Until this date, while almost all fabrication efforts have been directed toward multiwalled CNT (MWCNT) interconnects, there is a lack of MWCNT modeling work. This paper presents, for the first time, a detailed investigation of MWCNT-based interconnect performance. A compact equivalent circuit model of MWCNTs is presented for the first time, and the performance of MWCNT interconnects is evaluated and compared against traditional Cu interconnects, as well as Single-Walled CNT (SWCNT)-based interconnects, at different interconnect levels (local, intermediate, and global) for future technology nodes. It is shown that at the intermediate and global levels, MWCNT interconnects can achieve smaller signal delay than that of Cu interconnects, and the improvements become more significant with technology scaling and increasing wire lengths. At 1000- global or 500- intermediate level interconnects, the delay of MWCNT interconnects can reach as low as 15% of Cu interconnect delay. It is also shown that in order for SWCNT bundles to outperform MWCNT interconnects, dense and high metallic-fraction SWCNT bundles are necessary. On the other hand, since MWCNTs are easier to fabricate with less concern about the chirality and density control, they can be attractive for immediate use as horizontal wires in VLSI, including local, intermediate, and global level interconnects.     

Interconnect design strategy: structures, repeaters and materialswith strategic system performance analysis (S2PAL) model

In this paper, we propose a novel methodology for scheming an interconnect strategy, such as what interconnect structure should be taken, how repeaters should be inserted, and when new metal or dielectric materials should be adopted. In the methodology, the strategic system performance analysis model is newly developed as a calculation model that predicts LSI operation frequency and chip size with electrical parameters of transistors and interconnects as well as circuit configuration. The analysis with the model indicates that interconnect delay overcomes circuit block cycle time at a specific length; D_c-cross. Here tentatively, interconnects shorter than D_c-cross are called local interconnects, and interconnects longer than that as global ones. The cross-sectional structures for local and global tiers are optimized separately. We also calculate global interconnect pitch and the chip size enlarged by the global interconnect pitch and the inserted repeaters, and then estimate the effectiveness of introducing new materials for interconnects and dielectrics     

Wafer-level package interconnect options

As integrated circuit technology enters the nanometer era, global interconnects are becoming a bottleneck for overall chip performance. In this paper, we show that wafer-level package interconnects are an effective alternative to conventional on-chip global wires. These interconnects behave as LC transmission lines and can be exploited for their near speed of light transmission and low attenuation characteristics. We compare performance measures such as bandwidth, bandwidth density, latency, and power consumption of the package-level transmission lines with conventional on-chip global interconnects for different International Technology Roadmap for Semiconductors (ITRS) technology nodes. Based on these results, we show that package-level interconnects are well suited for power demanding low-latency applications. We also analyze different interconnect options such as memory buses, long inter tile interconnects, clock, and power distribution.     

Modeling, Analysis, and Design of Graphene Nano-Ribbon Interconnects

Graphene nanoribbons (GNRs) are considered as a prospective interconnect material. A comprehensive conductance and delay analysis of GNR interconnects is presented in this paper. Using a simple tight-binding model and the linear response Landauer formula, the conductance model of GNR is derived. Several GNR structures are examined, and the conductance among them and other interconnect materials [e.g., copper (Cu), tungsten (W), and carbon nanotubes (CNTs)] is compared. The impact of different model parameters (i.e., bandgap, mean free path, Fermi level, and edge specularity) on the conductance is discussed. Both global and local GNR interconnect delays are analyzed using an RLC equivalent circuit model. Intercalation doping for multilayer GNRs is proposed, and it is shown that in order to match (or better) the performance of Cu or CNT bundles at either the global or local level, multiple zigzag-edged GNR layers along with proper intercalation doping must be used and near-specular nanoribbon edge should be achieved. However, intercalation-doped multilayer zigzag GNRs can have better performance than that of W, implying possible application as local interconnects in some cases. Thus, this paper identifies the on-chip interconnect domains where GNRs can be employed and provides valuable insights into the process technology development for GNR interconnects.      

Analysis and optimization of shielded RLC global interconnects for gigascale integration

Compact physical models are presented for on-chip double-sided shielded transmission lines, which are mainly used for long global interconnects where inductance effects should not be ignored. The models are then used to optimize the width and spacing of long global interconnects with repeater insertion. The impacts of increasing line width and spacing on various performance parameters such as delay, data-flux density, power dissipation and total repeater area are analysed. The product of data-flux density and reciprocal delay per unit length are defined as a figure of merit (FOM). By maximizing the FOM, the optimal width and spacing of shielded RLC global interconnects are obtained for various international technology roadmap for semiconductors (ITRS) technology nodes.     

Bus encoder design for reduced crosstalk,power and area in coupled VLSI interconnects

This research work presents a novel circuit for simultaneous reduction of power, crosstalk and area using bus encoding technique in RC modeled VLSI interconnect. Bus-invert method is used to reduce inter-wire coupling, which is actually responsible for crosstalk, delay and power dissipation in coupled interconnects. The proposed method focuses on simplified and improved encoder circuit for 4, 8 and 16 coupled lines. In past, the researchers developed encoders that usually focused on minimizing power dissipation and/or crosstalk, thereby paying heavy penalty in terms of chip area. However, the proposed encoder and decoder while significantly reducing crosstalk demonstrates an overall reduction of power dissipation by 68.76% through drastically limiting switching activity. Furthermore, while reducing the complexity, chip area and transistor count of the circuit is reduced by more than 57%.     

深亚微米多层互连的温度分布特性分析

本文从热扩散方程出发,推导了简单互连的温度分布解析表达式,采用65nm工艺参数,详细讨论了热扩散长度和介质层厚度对互连温度分布的影响;进一步给出了复杂多层互连的温度分布解析表达式并用于其特性模拟,结果显示全局互连的温升远大于半全局互连和局部互连的温升。     

A 32-Gb/s On-Chip Bus With Driver Pre-Emphasis Signaling

This paper describes a differential current-mode bus architecture based on driver pre-emphasis for on-chip global interconnects that achieves high-data rates while reducing bus power dissipation and improving signal delay latency. The 16-b bus core fabricated in 0.25- $mu$m complementary metal–oxide–semiconductor (CMOS) technology attains an aggregate signaling data rate of 32 Gb/s over 5–10-mm-long lossy interconnects. With a supply of 2.5 V, 25.5–48.7-mW power dissipation was measured for signal activity above 0.1, equivalent to 0.80–1.52 pJ/b. This work demonstrates a 15.0%–67.5% power reduction over a conventional single-ended voltage-mode static bus while reducing delay latency by 28.3% and peak current by 70%. The proposed bus architecture is robust against crosstalk noise and occupies comparable routing area to a reference static bus design.      

Interconnect limits on gigascale integration (GSI) in the 21stcentury

Twenty-first century opportunities for GSI will be governed in part by a hierarchy of physical limits on interconnects whose levels are codified as fundamental, material, device, circuit, and system. Fundamental limits are derived from the basic axioms of electromagnetic, communication, and thermodynamic theories, which immutably restrict interconnect performance, energy dissipation, and noise reduction. At the material level, the conductor resistivity increases substantially in sub-50-nm technology due to scattering mechanisms that are controlled by quantum mechanical phenomena and structural/morphological effects. At the device and circuit level, interconnect scaling significantly increases interconnect crosstalk and latency. Reverse scaling of global interconnects causes inductance to influence on-chip interconnect transients such that even with ideal return paths, mutual inductance increases crosstalk by up to 60% over that predicted by conventional RC models. At the system level, the number of metal levels explodes for highly connected 2-D logic megacells that double in size every two years such that by 2014 the number is significantly larger than ITRS projections. This result emphasizes that changes in design, technology, and architecture are needed to cope with the onslaught of wiring demands. One potential solution is 3-D integration of transistors, which is expected to significantly improve interconnect performance. Increasing the number of active layers, including the use of separate layers for repeaters, and optimizing the wiring network, yields an improvement in interconnect performance of up to 145% at the 50-nm node     

Vertically integrated SOI circuits for low-power and high-performance applications

Vertical integration offers numerous advantages over conventional structures. By stacking multiple-material layers to form double gate transistors and by stacking multiple device layers to form multidevice-layer integration, vertical integration can emerge as the technology of choice for low-power and high-performance integration. In this paper, we demonstrate that the vertical integration can achieve better circuit performance and power dissipation due to improved device characteristics and reduced interconnect complexity and delay. The structures of vertically integrated double gate (DG) silicon-on-insulator (SOI) devices and circuits, and corresponding multidevice-layer (3-D) SOI circuits are presented; a general double-gate SOI model is provided for the study of symmetric and asymmetric SOI CMOS circuits; circuit speed, power dissipation of double-gate dynamic threshold (DGDT) SOI circuits are investigated and compared to single gate (SG) SOI circuits; potential 3-D SOI circuits are laid out. Chip area, layout complexity, process cost, and impact on circuit performance are studied. Results show that DGDT SOI CMOS circuits provide the best power-delay product, which makes them very attractive for low-voltage low-power applications. Multidevice-layer integration achieves performance improvement by shortening the interconnects. Results indicate that up to 40% of interconnect performance improvements can be expected for a 4-device-layer integration.     

High-speed and low-power repeater for VLSI interconnects

This paper proposes a repeater for boosting the speed of interconnects with low power dissipation. We have designed and implemented at 45 and 32 nm technology nodes. Delay and power dissipation performances are analyzed for various voltage levels at these technology nodes using Spice simulations. A significant reduction in delay and power dissipation are observed compared to a conventional repeater. The results show that the proposed high-speed low-power repeater has a reduced delay for higher load capacitance. The proposed repeater is also compared with LPTG CMOS repeater, and the results shows that the proposed repeater has reduced delay. The proposed repeater can be suitable for high-speed global interconnects and has the capacity to drive large loads.     

Propagation delay and power dissipation for different aspect ratio of single-walled carbon nanotube bundled TSV

Through-silicon vias (TSVs) have provided an attractive solution for three-dimensional (3D) integrated devices and circuit technologies with reduced parasitic losses and power dissipation, higher input-output (I/O) density and improved system performance. This paper investigates the propagation delay and average power dissipation of single-walled carbon nanotube bundled TSVs having different via radius and height. Depending on the physical configuration, a comprehensive and accurate analytical model of CNT bundled TSV is employed to represent the via (vertical interconnect access) line of a driver-TSV-load (DTL) system. The via radius and height are used to estimate the bundle aspect ratio (AR) and the cross-sectional area. For a fixed via height, the delay and the power dissipation are reduced up to 96.2% using a SWCNT bundled TSV with AR = 300 : 1 in comparison to AR = 6 : 1.     

Optimal global interconnects for GSI

Performance of a high-speed chip is largely affected by both latency and bandwidth of global interconnects, which connect different macrocells. Therefore, one of the important goals is to design high-bandwidth and fast buses that connect a processor and its on-chip cache memory or link different processors within a multiprocessor chip. In this paper, the width of global interconnects is optimized to achieve a large "data-flux density" and a small latency simultaneously. Data-flux density is the product of interconnect bandwidth and reciprocal wire pitch, which represents the number of bits per second that can be transferred across a unit-length bisectional line. The optimal wire width, which maximizes the product of data-flux density and reciprocal latency, is independent of interconnect length and can be used for all global interconnects. It is rigorously proved that the optimal wire width is the width that results in a delay that is 33% larger than the time-of-flight (ToF). Using the optimal wire width decreases latency, energy dissipation, and repeater area considerably, compared to a sub-optimal wire width (e.g., 42% smaller latency, 30% smaller energy-per-bit, and 84% smaller repeater area compared with the W/sub opt//2 case) at the cost of a small decrease in data-flux density (e.g., 14% smaller compared with W/sub opt//2 case). A super-optimal wire width, however, causes a slight decrease in latency (e.g., 14% for 2W/sub opt/) at the cost of a large decrease in data-flux density (e.g., 35% for 2W/sub opt/).