ESSDERC/ESSCIRC Workshop
cmos variability research in europe: from atomic scale to circuits and systems

Edinburgh International Conference Centre
19th September 2008

Variability in FinFETs : Experiments and model


Abdelkarim Mercha

Figure 1: Top: Comparison of the experimental (open symbols) and predicted (plain lines) normalized mismatch in the drain current versus gate overdrive, for n-type and p-type FinFETs. Bottom: Error between the experimental and modeled ID mismatch.
As Moore’s law is leading the CMOS technology into the decananometer regime, the behavior of the MOSFET deviates from that of an ideal switch. The SCEs are the main source of this deviation, resulting in an increasingly difficult control over the scaled channels. Besides the SCEs, variations in various process parameters and their fluctuations also become higher as the channel area shrinks. These give rise to variability and mismatch in the transistor characteristics, affecting both the product performance and yield. The introduction of alternative multiple gate transistor architectures, such as SOI FinFET, can significantly improve transistor performance by a stronger coupling to the channel, and lower channel doping. Measurements and modeling of the mismatch in drain current, threshold voltage, and current factor between FinFET pairs are presented and developed for most advanced technology options [1-3]. An overview comparing FinFET performances with state of the art technologies and the ITRS is given. Furthermore, extracted mismatch parameters are used to assessing the impact of mismatch on FinFET circuit design for both analog and SRAM applications.

[1] “Stochastic Matching Properties of FinFETs”, C.Gustin, A. Mercha et al., EDL, 27, 10, 846-848, 2006.
[2] “Impact of advanced process modules and device architectures on the matching performance of (sub-)45nm CMOS”, C.Gustin, A. Mercha et al., VLSI-TSA 2007. 23-25 April 2007
[3] “Impact of Stochastic Mismatch on Measured SRAM Performance of FinFETs with Resist/Spacer-Defined Fins: Role of Line-Edge-Roughness” A. Dixit,; Anil, K. G.; et al. proceedings IEDM 2006, 11-13
[4] “Analog design challenges and trade-offs using emerging materials and devices” M.Fulde, A. Mercha et al. Proceedings ESSDERC 2007 123-126
modeling summer school semiconductors Semiconductors semiconductor devices education training microelectronics industry medici TMA suprem workbench Synopsys Silvaco device modeling device simulation semiconductor simulation process simulation diffusion ion implantation impurities oxidation furnace finite element industrial services finite elements calibration design semiconductor research University of Glasgow electronics electrical engineering courses MOSFET CMOS transistor BJT diode doping doping profile electrons holes potential concentration fabrication silicon Si gallium arsenide GaAs silicon germanium III-V SiGe quantum mechanics transport band structures IWCE IEDM SISPAD ESSDERC density gradient taurus monte-carlo monte carlo thin body strained silicon CMOS SiNano device physics atomistic SOI greens functions green post doctoral academic glasgow tutorial summerschool simulation