Search

The optimal design of repeaters for current-mode (CM) nano-interconnects is conducted and compared with that for voltage-mode (VM) ones. Both Cu and carbon nanotube (CNT) nano-interconnects are considered and studied, with the contact resistance treated appropriately. In comparison with the VM ones, the CM nano-interconnects can not only demonstrate better electrical performance, but also require fewer repeaters, thereby saving more valuable chip area. In the delay-optimal repeater design of multi-walled CNT interconnects, for instance, the optimal number of repeaters can be reduced by 120% with the implementation of CM signalling scheme.

N-substituted bis(diphenylphosphanyl)amine RN(PPh 2 ) 2 (PNP; R = CH 2 CH 2 CH 2 Si(OEt) 3 ) and its mononuclear nickel(II) ethanedithiolate complexe RN(PPh 2 ) 2 Ni(SCH 2 CH 2 S) (PNS; R = CH 2 CH 2 CH 2 Si(OEt) 3 ) were prepared and used as multi-element flame retardants for epoxy resin (EP). Results of thermogravimetric analysis (TGA) revealed that the incorporation of PNP or PNS restrained the decomposition of EP and also improved the stability of char residues. Moreover, results of limited oxygen index (LOI), vertical burning test (UL-94), microscale combustion calorimeter (MCC), and cone calorimeter tests proved that both PNP and PNS endowed EP with good flame retardancy. Different chemical structures and action mechanisms of PNP and PNS resulted in different degrees of flame retardancy for EP. Although EP/PNP samples caused a little higher LOI values than those of EP/PNS ones, the incorporation of 7 wt% PNS endowed EP with a lower peak of heat release rate (PHRR) and total smoke production (TSP), thereby indicating better flame-retardant and smoke suppression effects. Furthermore, thermogravimetry-Fourier transform infrared spectroscopy (TG-FTIR) and pyrolysis gas chromatography/mass spectrometry (Py-GC/MS) were employed to analyze the pyrolysis behavior of EPs and flame retardants. The morphologies of char layers for EPs were examined by scanning electron microscopy (SEM) and energy dispersive X-ray spectroscopy (EDX). In comparison to PNP, PNS promoted the formation of a stronger protective char barrier comprising by P/Si/S/Ni elements and also restrained the effects of flammable gases on EP.

Optimal repeater designs are performed for Cu and carbon nanotube (CNT)-based nanointerconnects to reduce the delay and power dissipation. The effects of inductance and metal-CNT contact resistance are treated appropriately. In this paper, the circuit parameters are calculated analytically, while they can be extracted experimentally for a specific foundry at a specific technology node. The particle swarm optimization (PSO) technique is employed to numerically calculate the optimal repeater size and the optimal number of repeaters in the Cu and CNT-based nanointerconnects. The results are verified against the analytical and genetic algorithm results. To facilitate CAD design, the machine-learning neural network (NN) is adopted. The data obtained using the PSO algorithm are used to train the NN and the feasibility of the NN is investigated and validated.

In this paper, the optimal repeater number and size are analyzed for multi-walled carbon nanotube interconnects by using the particle swarm optimization (PSO) algorithm. Genetic algorithm (GA) is also used to verify the corresponding results. Further, the neural network (NN) is trained to facilitate the EDA process. It is found that the computational time can be dramatically reduced with the implementation of NN.

Closed-form expressions for the optimized number and size of repeaters in multi-walled carbon nanotube (MWCNT) interconnects are presented. The contact resistance and inductive effects are taken into account. It is found that the propagation delay of MWCNT interconnects can be reduced effectively by inserting repeaters. However, the contact resistance has a significant influence on the optimized number and size of repeaters. Moreover, it is found that both the optimal number of repeaters and the minimum propagation delay are kept almost unchanged with the variation of carbon nanotube (CNT) kinetic inductance. The optimal number of repeaters in the MWCNT interconnect is much smaller than that in its Cu counterpart, thus saving chip area and power consumption.