Efficient Cache Exploration Method for a Tiled Chip Multiprocessor

Past studies use deterministic models to evaluate optimal cache configuration or to explore its design space. However, with the increasing number of components present on a chip multiprocessor (CMP), deterministic approaches do not scale well. Hence, we apply probabilistic genetic algorithms (GA) to determine a near-optimal cache configuration for a sixteen tiled CMP. We propose and implement a faster trace based approach to estimate fitness of a chromosome. It shows up-to 218x simulation speedup over the cycle-accurate architectural simulation. Our methodology can be applied to solve other cache optimization problems such as design space exploration of cache and its partitioning among applications/ virtual machines.

An approximately H-1-optimal Petrov-Galerkin meshfree method: application to computation of scattered light for optical tomography

Nearly pollution-free solutions of the Helmholtz equation for k-values corresponding to visible light are demonstrated and verified through experimentally measured forward scattered intensity from an optical fiber. Numerically accurate solutions are, in particular, obtained through a novel reformulation of the H-1 optimal Petrov-Galerkin weak form of the Helmholtz equation. Specifically, within a globally smooth polynomial reproducing framework, the compact and smooth test functions are so designed that their normal derivatives are zero everywhere on the local boundaries of their compact supports. This circumvents the need for a priori knowledge of the true solution on the support boundary and relieves the weak form of any jump boundary terms. For numerical demonstration of the above formulation, we used a multimode optical fiber in an index matching liquid as the object. The scattered intensity and its normal derivative are computed from the scattered field obtained by solving the Helmholtz equation, using the new formulation and the conventional finite element method. By comparing the results with the experimentally measured scattered intensity, the stability of the solution through the new formulation is demonstrated and its closeness to the experimental measurements verified.

On the mechanism of metal nanoparticle synthesis in the brust-schiffrin method

Brust-Schiffrin synthesis (BSS) of metal nanoparticles has emerged as a major breakthrough in the field for its ability to produce highly stable thiol functionalized nanoparticles. In this work, we use a detailed population balance model to conclude that particle formation in BSS is controlled by a new synthesis route: continuous nucleation, growth, and capping of particles throughout the synthesis process. The new mechanism, quite different from the others known in the literature (classical LaMer mechanism, sequential nucleation-growth-capping, and thermodynamic mechanism), successfully explains key features of BSS, including size tuning by varying the amount of capping agent instead of the widely used approach of varying the amount of reducing agent. The new mechanism captures a large body of experimental observations quantitatively, including size tuning and only a marginal effect of the parameters otherwise known to affect particle synthesis sensitively. The new mechanism predicts that, in a constant synthesis environment, continuous nucleation-growth-capping mechanism leads to complete capping of particles (no more growth) at the same size, while the new ones are born continuously, in principle leading to synthesis of more monodisperse particles. This prediction is validated through new experimental measurements.

Hybrid aggregated-vector algorithm for efficient parallelization of fast multipole method

An efficient parallelization algorithm for the Fast Multipole Method which aims to alleviate the parallelization bottleneck arising from lower job-count closer to root levels is presented. An electrostatic problem of 12 million non-uniformly distributed mesh elements is solved with 80-85% parallel efficiency in matrix setup and matrix-vector product using 60GB and 16 threads on shared memory architecture.

Partial delamination modeling in composite beams using a finite element method

A new method of modeling partial delamination in composite beams is proposed and implemented using the finite element method. Homogenized cross-sectional stiffness of the delaminated beam is obtained by the proposed analytical technique, including extension-bending, extension-twist and torsion-bending coupling terms, and hence can be used with an existing finite element method. A two noded C1 type Timoshenko beam element with 4 degrees of freedom per node for dynamic analysis of beams is implemented. The results for different delamination scenarios and beams subjected to different boundary conditions are validated with available experimental results in the literature and/or with the 3D finite element simulation using COMSOL. Results of the first torsional mode frequency for the partially delaminated beam are validated with the COMSOL results. The key point of the proposed model is that partial delamination in beams can be analyzed using a beam model, rather than using 3D or plate models. (c) 2013 Elsevier B.V. All rights reserved.

The effect of Sb on the electrical and magnetic properties of Ni-Zn ferrites prepared by sol-gel autocombustion method

Polycrystalline Ni-Zn ferrites with a well-defined composition of Ni0.4Zn0.6Fe2-xSbxO4 synthesized using sol-gel method. Morphological characterizations on the prepared samples were performed by high resolution transmission electron and field emission scanning electron microscopy. The powders were densified using microwave sintering method. The room temperature complex permittivity (epsilon' and epsilon aEuro(3)) and permeability (mu' and mu aEuro(3)) were measured over a wide frequency range from 1 MHz-1.8 GHz. The real part of permittivity varies as `x' concentration increases and the resonance frequency was observed at much higher frequencies and there is a significant decrease in the loss factor (tan delta). The electrical resistivity and permeability of NiZn ferrites increased with an increase of Sb content. As the concentration of `x' increases from 0 to 0.08 the saturation magnetisation decreases. The saturation magnetization (M-s) a parts per thousand aEuro parts per thousand 52.211 A.m(2)/Kg for x = 0 at room temperature. The room temperature electro paramagnetic resonance (EPR) were studied.

A study of the speed and the accuracy of the Boundary Element Method as applied to the computational simulation of biological organs

In this work, first a Fortran code is developed for three dimensional linear elastostatics using constant boundary elements; the code is based on a MATLAB code developed by the author earlier. Next, the code is parallelized using BLACS, MPI, and ScaLAPACK. Later, the parallelized code is used to demonstrate the usefulness of the Boundary Element Method (BEM) as applied to the realtime computational simulation of biological organs, while focusing on the speed and accuracy offered by BEM. A computer cluster is used in this part of the work. The commercial software package ANSYS is used to obtain the `exact' solution against which the solution from BEM is compared; analytical solutions, wherever available, are also used to establish the accuracy of BEM. A pig liver is the biological organ considered. Next, instead of the computer cluster, a Graphics Processing Unit (GPU) is used as the parallel hardware. Results indicate that BEM is an interesting choice for the simulation of biological organs. Although the use of BEM for the simulation of biological organs is not new, the results presented in the present study are not found elsewhere in the literature. Also, a serial MATLAB code, and both serial and parallel versions of a Fortran code, which can solve three dimensional (3D) linear elastostatic problems using constant boundary elements, are provided as supplementary files that can be freely downloaded.

An asymptotically optimal push-pull method for multicasting over a random network

We consider multicast flow problems where either all of the nodes or only a subset of the nodes may be in session. Traffic from each node in the session has to be sent to every other node in the session. If the session does not consist of all the nodes, the remaining nodes act as relays. The nodes are connected by undirected edges whose capacities are independent and identically distributed random variables. We study the asymptotics of the capacity region (with network coding) in the limit of a large number of nodes, and show that the normalized sum rate converges to a constant almost surely. We then provide a decentralized push-pull algorithm that asymptotically achieves this normalized sum rate.