Search

We conducted a performance evaluation of a production supercomputer based on the Intel Xeon Processor E5-2680v4 (Broadwell), a fourteen-core die shrink of the Haswell architecture and compared and contrasted it with two earlier Xeons: the E5-2680v3 (Haswell), a twelve-core implementation of the fourth-generation Haswell architecture and the E5-2680v2 (Ivy Bridge), a ten-core die shrink of the third-generation Sandy Bridge architecture. The new features of Broadwell architecture are evaluated and compared with Ivy Bridge using several low-level benchmarks and five full-scale production quality scientific and engineering applications. A model to predict the performance of HPCG, Overflow and Cart3D is presented for two architectures. In addition, we evaluate the effects of degrading base core frequency to AVX frequency while using AVX2 instructions and of the Turbo Boost feature.

We present a performance evaluation conducted on a production supercomputer of the Intel Xeon Processor E5-2680v3, a twelve-core implementation of the fourth-generation Haswell architecture, and compare it with Intel Xeon Processor E5-2680v2, an Ivy Bridge implementation of the third-generation Sandy Bridge architecture. Several new architectural features have been incorporated in Haswell including improvements in all levels of the memory hierarchy as well as improvements to vector instructions and power management. We critically evaluate these new features of Haswell and compare with Ivy Bridge using several low-level benchmarks including subset of HPCC, HPCG and four full-scale scientific and engineering applications. We also present a model to predict the performance of HPCG and Cart3D within 5%, and Overflow within 10% accuracy.

The syndrome of familial hypoparathyroidism, sensorineural deafness, and renal dysplasia (HDR syndrome) is inherited as an autosomal dominant trait, caused by haploinsufficiency of the GATA3 gene in chromosome 10p. Although first described years ago, but the disease is considered to be very rare. Patients usually present with hypocalcemia, tetany, or afebrile convulsions at any age. Hearing loss is usually bilateral, range from mild to profound impairment. Renal disease includes dysplasia, hypoplasia or aplasia.

Resource management is an important component of a grid computing infrastructure. The scalability and adaptability of such systems are two key challenges that must be addressed. In this work an agent-based resource management system, ARMS, is implemented for grid computing. ARMS utilises the performance prediction techniques of the PACE toolkit to provide quantitative data regarding the performance of complex applications running on a local grid resource. At the meta-level, a hierarchy of homogeneous agents are used to provide a scalable and adaptable abstraction of the system architecture. Each agent is able to cooperate with other agents and thereby provide service advertisement and discovery for the scheduling of applications that need to utilise grid resources. A case study with corresponding experimental results is included to demonstrate the efficiency of the resource management and scheduling system.

A suite of scalable atomistic simulation programs has been developed for materials research based on space-time multiresolution algorithms. Design and analysis of parallel algorithms are presented for molecular dynamics (MD) simulations and quantum-mechanical (QM) calculations based on the density functional theory. Performance tests have been carried out on 1,088-processor Cray T3E and 1,280-processor IBM SP3 computers. The linear-scaling algorithms have enabled 6.44-billion-atom MD and 111,000-atom QM calculations on 1,024 SP3 processors with parallel efficiency well over 90%. production-quality programs also feature wavelet-based computational-space decomposition for adaptive load balancing, spacefilling-curve-based adaptive data compression with user-defined error bound for scalable I/O, and octree-based fast visibility culling for immersive and interactive visualization of massive simulation data.

In this paper, we present a detailed simulation study of the influence of quantum mechanical effects in the inversion layer on random dopant induced threshold voltage fluctuations and lowering in sub-100 mn MOSFETs. The simulations have been performed using a three-dimensional (3-D) implementation of the density gradient (DG) formalism incorporated in our established 3-D atomistic simulation approach. This results in a self-consistent 3-D quantum mechanical picture, which implies not only the vertical inversion layer quantization but also the lateral confinement effects related to current filamentation in the "valleys" of the random potential fluctuations. We have shown that the net result of including quantum mechanical effects, while considering statistical dopant fluctuations, is an increase in both threshold voltage fluctuations and lowering. At the same time, the random dopant induced threshold voltage lowering partially compensates for the quantum mechanical threshold voltage shift in aggressively scaled MOSFETs with ultrathin gate oxides.

We use the density gradient (DG) simulation approach to study, in 3D, the effect of local oxide thickness fluctuations on the threshold voltage of decanano MOSFETs in a statistical manner. A description of the reconstruction procedure for the random 2D surfaces representing the ‘atomistic’ Si–SiO 2 interface variations is presented. The procedure is based on power spectrum synthesis in the Fourier domain and can include either Gaussian or exponential spectra. The simulations show that threshold voltage variations induced by oxide thickness fluctuation become significant when the gate length of the devices become comparable to the correlation length of the fluctuations. The extent of quantum corrections in the simulations with respect to the classical case and the dependence of threshold variations on the oxide thickness are examined.

In this paper, we investigate various aspects of the polysilicon gate influence on the random dopant induced threshold voltage fluctuations in sub-100 nm MOSFETs with ultrathin gate oxides. The study is done by using an efficient statistical three-dimensional (3D) “atomistic” simulation technique. MOSFETs with uniform channel doping and with low doped epitaxial channels have been investigated. The simulations reveal that even in devices with a single crystal gate the gate depletion and the random dopants in it are responsible for a substantial fraction of the threshold voltage fluctuations when the gate oxide is scaled to thickness in the range of 1-2 nm. Simulation experiments have been used in order to separate the enhancement in the threshold voltage fluctuations due to an effective increase in the oxide thickness associated with the gate depletion from the direct influence of the random dopants in the gate depletion layer. The results of the experiments show that the both factors contribute to the enhancement of the threshold voltage fluctuations, but the effective increase in the oxide thickness has a dominant effect in the investigated range of devices. Simulations illustrating the effect of the polysilicon grain boundaries on the threshold voltage variation are also presented.

The possibility of nanoscale etching and indentation of Si(001)(2x1) surface by (8,0) and (10,10) carbon nanotube tips is demonstrated, for the first time, by classical molecular dynamics simulations employing Tersoff's many-body potential for a mixed C/Si/Ge system. In the nanotube tip barely touching the surface scenario atomistic etching is observed, where as in the nanoindentation scenario nanotube tip penetrates the surface without much hindrance. The results are explained in terms of the relative strength of C-C, C-Si, and Si-Si bonds.

A detailed three-dimensional (3-D) statistical "atomistic" simulation study of fluctuation-resistant sub 0.1-/spl mu/m MOSFET architectures with epitaxial channels and delta doping is presented. The need for enhancing the fluctuation resistance of the sub-0.1-/spl mu/m generation transistors is highlighted by presenting summarized results from atomistic simulations of a wide range of conventional devices with uniformly doped channels. According to our atomistic results, the doping concentration dependence of the random dopant-induced threshold voltage fluctuations in conventional devices is stronger than the analytically predicted fourth-root dependence. As a result of this, the scaling of such devices will be restricted by the "intrinsic" random dopant-induced fluctuations earlier than anticipated. Our atomistic simulations confirm that the introduction of a thin epitaxial layer in the MOSFET's channel can efficiently suppress the random dopant-induced threshold voltage fluctuations in sub-0.1-/spl mu/m devices. For the first time, we observe an "anomalous" reduction in the threshold voltage fluctuations with an increase in the doping concentration behind the epitaxial channel, which we attribute to screening effects. Also, for the first time we study the effect of a delta doping, positioned behind the epitaxial layer, on the intrinsic threshold voltage fluctuations. Above a certain thickness of epitaxial layer, we observe a pronounced anomalous decrease in the threshold voltage fluctuation with the increase of the delta doping. This phenomenon, which is also associated with screening, enhances the importance of the delta doping in the design of properly scaled fluctuation-resistant sub-0.1-/spl mu/m MOSFET's.