Automatic Compilation of MATLAB Programs for Synergistic Execution on Heterogeneous Processors

MATLAB is an array language, initially popular for rapid prototyping, but is now being increasingly used to develop production code for numerical and scientific applications. Typical MATLAB programs have abundant data parallelism. These programs also have control flow dominated scalar regions that have an impact on the program's execution time. Today's computer systems have tremendous computing power in the form of traditional CPU cores and throughput oriented accelerators such as graphics processing units(GPUs). Thus, an approach that maps the control flow dominated regions to the CPU and the data parallel regions to the GPU can significantly improve program performance. In this paper, we present the design and implementation of MEGHA, a compiler that automatically compiles MATLAB programs to enable synergistic execution on heterogeneous processors. Our solution is fully automated and does not require programmer input for identifying data parallel regions. We propose a set of compiler optimizations tailored for MATLAB. Our compiler identifies data parallel regions of the program and composes them into kernels. The problem of combining statements into kernels is formulated as a constrained graph clustering problem. Heuristics are presented to map identified kernels to either the CPU or GPU so that kernel execution on the CPU and the GPU happens synergistically and the amount of data transfer needed is minimized. In order to ensure required data movement for dependencies across basic blocks, we propose a data flow analysis and edge splitting strategy. Thus our compiler automatically handles composition of kernels, mapping of kernels to CPU and GPU, scheduling and insertion of required data transfer. The proposed compiler was implemented and experimental evaluation using a set of MATLAB benchmarks shows that our approach achieves a geometric mean speedup of 19.8X for data parallel benchmarks over native execution of MATLAB.

Thermal degradation kinetics of poly(trimethylol propane triacrylate)/poly(hexane diol diacrylate) interpenetrating polymer network

Interpenetrating polymer networks (IPNs) of trimethylol propane triacrylate (TMPTA) and 1,6-hexane diol diacrylate (HDDA) at different weight ratios were synthesized. Temperature modulated differential scanning calorimetry (TMDSC) was used to determine whether the formation resulted in a copolymer or interpenetrating polymer network (IPN). These polymers are used as binders for microstereolithography (MSL) based ceramic microfabrication. The kinetics of thermal degradation of these polymers are important to optimize the debinding process for fabricating 3D shaped ceramic objects by MSL based rapid prototyping technique. Therefore, thermal and thermo-oxidative degradation of these IPNs have been studied by dynamic and isothermal thermogravimetry (TGA). Non-isothermal model-free kinetic methods have been adopted (isoconversional differential and KAS) to calculate the apparent activation energy (E a) as a function of conversion (α) in N 2 and air. The degradation of these polymers in N 2 atmosphere occurs via two mechanisms. Chain end scission plays a dominant role at lower temperature while the kinetics is governed by random chain scission at higher temperature. Oxidative degradation shows multiple degradation steps having higher activation energy than in N 2. Isothermal degradation was also carried out to predict the reaction model which is found to be decelerating. It was shown that the degradation of PTMPTA follows a contracting sphere reaction model in N 2. However, as the HDDA content increases in the IPNs, the degradation reaction follows Avrami-Erofeev model and diffusion governed mechanisms. The intermediate IPN compositions show both type of mechanism. Based on the above study, debinding strategy for MSL based microfabricated ceramic structure has been proposed. © 2012 Elsevier B.V.

Automatic compilation of MATLAB programs for synergistic execution on heterogeneous processors

MATLAB is an array language, initially popular for rapid prototyping, but is now being increasingly used to develop production code for numerical and scientific applications. Typical MATLAB programs have abundant data parallelism. These programs also have control flow dominated scalar regions that have an impact on the program's execution time. Today's computer systems have tremendous computing power in the form of traditional CPU cores and throughput oriented accelerators such as graphics processing units(GPUs). Thus, an approach that maps the control flow dominated regions to the CPU and the data parallel regions to the GPU can significantly improve program performance. In this paper, we present the design and implementation of MEGHA, a compiler that automatically compiles MATLAB programs to enable synergistic execution on heterogeneous processors. Our solution is fully automated and does not require programmer input for identifying data parallel regions. We propose a set of compiler optimizations tailored for MATLAB. Our compiler identifies data parallel regions of the program and composes them into kernels. The problem of combining statements into kernels is formulated as a constrained graph clustering problem. Heuristics are presented to map identified kernels to either the CPU or GPU so that kernel execution on the CPU and the GPU happens synergistically and the amount of data transfer needed is minimized. In order to ensure required data movement for dependencies across basic blocks, we propose a data flow analysis and edge splitting strategy. Thus our compiler automatically handles composition of kernels, mapping of kernels to CPU and GPU, scheduling and insertion of required data transfer. The proposed compiler was implemented and experimental evaluation using a set of MATLAB benchmarks shows that our approach achieves a geometric mean speedup of 19.8X for data parallel benchmarks over native execution of MATLAB.

Optimization of rheological properties of photopolymerizable alumina suspensions for ceramic microstereolithography

Microstereolithography (MSL) is a rapid prototyping technique to fabricate complex three-dimensional (3D) structure in the microdomain involving different materials such as polymers and ceramics. The present effort is to fabricate microdimensional ceramics by the MSL system from a non-aqueous colloidal slurry of alumina. This slurry predominantly consists of two phases i.e. sub-micrometer solid alumina particles and non-aqueous reactive difunctional and trifunctional acrylates with inert diluent. The first part of the work involves the study of the stability and viscosity of the slurry using different concentrations of trioctyl phosphine oxide (TOPO) as a dispersant. Based on the optimization, the highest achievable solid loadings of alumina has been determined for this particular colloidal suspension. The second part of the study highlights the fabrication of several micro-dimensional alumina structures by the MSL system. (C) 2013 Elsevier Ltd and Techna Group S.r.l. All rights reserved.

Conceptual design of three-dimensional scaffolds of powder-based materials for bone tissue engineering applications

Purpose - The purpose of this paper is to investigate the possibility to construct tissue-engineered bone repair scaffolds with pore size distributions using rapid prototyping techniques. Design/methodology/approach - The fabrication of porous scaffolds with complex porous architectures represents a major challenge in tissue engineering and the design aspects to mimic complex pore shape as well as spatial distribution of pore sizes of natural hard tissue remain unexplored. In this context, this work aims to evaluate the three-dimensional printing process to study its potential for scaffold fabrication as well as some innovative design of homogeneously porous or gradient porous scaffolds is described and such design has wider implication in the field of bone tissue engineering. Findings - The present work discusses biomedically relevant various design strategies with spatial/radial gradient in pore sizes as well as with different pore sizes and with different pore geometries. Originality/value - One of the important implications of the proposed novel design scheme would be the development of porous bioactive/biodegradable composites with gradient pore size, porosity, composition and with spatially distributed biochemical stimuli so that stem cells loaded into scaffolds would develop into complex tissues such as those at the bone-cartilage interface.

Growth and electron effective mass measurements of strained Si and Si0.94Ge0.06 on relaxed Si0.62Ge0.38 buffers grown by rapid thermal chemical vapor deposition

Abstract: We report the growth and the electron cyclotron resonance measurements of n-type Si/Si0.62Ge0.38 and Si0.94Ge0.06/Si0.62Ge0.38 modulation-doped heterostructures grown by rapid thermal chemical vapor deposition. The strained Si and Si0.94Ge0.06 channels were grown on relaxed Si0.62Ge0.38 buffer layers, which consist of 0.6 mu m uniform Si0.62Ge0.38 layers and 0.5 mu m compositionally graded relaxed SiGe layers from 0 to 38% Ge. The buffer layers were annealed at 800 degrees C for 1 h to obtain complete relaxation. A 75 Angstrom Si(SiGe) channel with a 100 Angstrom spacer and a 300 Angstrom 2 X 10(19) cm(-3) n-type supply layer was grown on the top of the buffer layers. The cross-sectional transmission electron microscope reveals that the dense dislocation network is confined to the buffer layer, and relatively few dislocations terminate on the surface. The plan-view image indicates the threading dislocation density is about 4 X 10(6) cm(-2). The far-infrared measurements of electron cyclotron resonance were performed at 4 K with the magnetic field of 4-8 T. The effective masses determined from the slope of the center frequency of the absorption peak versus applied magnetic field plot are 0.203m(0) and 0.193m(0) for the two dimensional electron gases in the Si and Si0.94Ge0.06 channels, respectively. The Si effective mass is very close to that of a two dimensional electron gas in an Si MOSFET (0.198m(0)). The electron effective mass of Si0.94Ge0.06 is reported for the first time and is about 5% lower than that of pure Si.

High microwave susceptibility of NaH2PO4 center dot 2H(2)O: Rapid synthesis of crystalline and glassy phosphates with NASICON-type chemistry

High microwave susceptibility of NaH2PO4 . 2H(2)O has been discovered, This hydrated acid phosphate of sodium can be heated upto 1000 K or more when exposed to 2.45 GHz microwaves. Using this, a novel microwave-assisted preparation of a number of important crystalline and glassy materials with NASICON-type chemistry has been accomplished in less than 8 min which is only a fraction of the time required for conventional synthetic procedures, The present single-shot approach to the preparation of phosphates is attractive in terms of its simplicity, rapidity, and general applicability, A ''step-ladder'' heating mechanism has been proposed to account for the high microwave absorbing ability of NaH2PO4 . 2H(2)O.

A rapid and highly selective method for the estimation of pyro-,tri- and orthophosphates

A rapid, highly selective and simple method has been developed for the quantitative determination of pyro-, tri- and orthophosphates. The method is based on the formation of a solid complex of bis(ethylenediamine)cobalt(III) species with pyrophosphate at pH 4.2-4.3, with triphosphate at pH 2.0-2.1 and with orthophosphate at pH 8.2-8.6. The proposed method for pyro- and triphosphates differs from the available method, which is based on the formation of an adduct with tris(ethylenediamine)cobalt(III) species. The complexes have the composition [Co(en)(2)HP2O7]4H(2)O and [Co(en)(2)H2P3O10]2H(2)O, respectively. The precipitation is instantaneous and quantitative under the recommended optimum conditions giving 99.5% gravimetric yield in both cases. There is no interferences from orthophosphate, trimetaphosphate and pyrophosphate species in the triphosphate estimation up to 5% of each component. The efficacy of the method has been established by determining pyrophosphate and triphosphate contents in various matrices. In the case of orthophosphate, the proposed method differs from the available methods such as ammonium phosphomolybdate, vanadophosphomolybdate and quinoline phosphomolybdate, which are based on the formation of a precipitate, followed by either titrimetry or gravimetry. The precipitation is instantaneous and the method is simple. Under the recommended pH and other reaction conditions, gravimetric yields of 99.6-100% are obtainable. The method is applicable to orthophosphoric acid and a variety of phosphate salts.

Rapid Synthesis And Properties Of Fevo4, Alvo4, Yvo4, And Eu3+-Doped Yvo4

Fine particle FeVO4, AIVO4, YVO4 and Yo.95Eu0.05VO4 have been prepared by the combustion of aqueous solutions containing corresponding metal nitrate, ammonium metavanadate, ammonium nitrate and 3-methyl-5-pyrazolone.The solutions containing the redox mixtures, when rapidly heated at 370 °C, ignite and undergo self-propagating,gas-producing, exothermic reactions to yield fine particle metal vanadates. Formation of crystalline vanadates was confirmed by powder X-ray diffraction patterns,27A1 nuclear magnetic resonance, IR and fluorescence spectra.

An avidin-biotin microELISA for rapid measurement of total and allergen-specific human IgE

his paper describes an improved microtiter solid-phase enzyme immunoassay for the determination of total and allergen-specific human IgE. This assay technique is unique in its use of the avidin-biotin interaction to increase sensitivity. The avidin-biotin microtiter enzyme-linked immunosorbant assay (AB-microELISA) was performed in polyvinyl chloride microtiter plates using biotinylated anti-IgE and horseradish peroxidase (HRP)-avidin conjugate. This AB-microELISA technique enabled the quantitation of human serum IgE in the range of 0.1–5 ng/ml (10–500 pg/test) in less than 3 h. Total serum IgE, whether measured by the AB-microELISA or the paper radioimmunosorbant test (PRIST) was similar (correlation coefficient, r = 0.92). Further, the presence or absence of positive skin tests to 7 specific allergens determined in serum donors generally agreed with the presence or absence of allergen-specific IgE in their sera as measured by the AB-microELISA. The quantity of short ragweed allergen-specific IgE as determined by the AB-microELISA agreed with values obtained by the radioimmunosorbant test (RAST) (correlation coefficient, r = 0.89). No significant interference by ragweed-specific IgG (blocking antibody) was observed in the quantitation of allergen-specific IgE. The AB-microELISA is not only rapid and inexpensive, but also more sensitive than other published ELISA procedures and comparable to solid-phase radioimmunoassays in the quantitation of total and allergen-specific IgE.

A rapid procedure for the isolation of spheroplasts from Mycobacterium smegmatis

Abstract is not available.

A rapid and sensitive method for the estimation of individual tRNA pools

A rapid and sensitive method is described to quantitatively compare tRNA pools for individual aminoacids in a single experiment. The procedure comprises of: (i) charging of total tRNA with a mixture of radiolabeled aminoacids, (ii) deacylation of the esterified tRNA with a volatile base and the recovery of the labeled aminoacid, (iii) derivatisation of the aminoacid with phenylisothiocyanate after mixing with excess of nonradioactive aminoacids, (iv) baseline separation of the phenylthiocarbamyl aminoacids by reverse phase high performance liquid chromatography monitored by A254nm and (v) quantitation of the radioactivity in individual aminoacid peaks. The radioactivity in the aminoacid peak corresponds to the quantity of the aminoacylated tRNA. The method has been successfully applied to quantitate the individual tRNA pools in the developing silk glands of Bombyx mori, a functionally adapted tissue which undergoes considerable variations in tRNA content. PSG, posterior silk gland; PITC, phenylisothiocyanate; DMAA, N,N-dimethyl-N-allylamine; APH, algal protein hydrolysate; ptc-, phenylthiocarbamyl; HPLC, high performance liquid chromatography.

The effect of rapid solidification on the order-disorder transformation in iron-based alloys

The effect of rapid solidification on the ordering reaction in Fe---Si and Fe---Al alloys has been reported. It is shown that rapid solidification can influence the ordering reaction in alloys with higher critical ordering temperatures. For ordering reactions at lower temperatures, the effect is similar to that of solid-state quenching. Different factors influencing the ordering reactions and domain structures during rapid solidification of iron-based alloys are discussed.

Synthetic studies towards bioactive frondosins: rapid framework access and diversity creation

A very concise and diversity-oriented approach to rapidly access frondosin-related frameworks from commercially available building blocks is outlined.

Rapid in situ processing for real-time holographic interferometry

Experiments are described which show that a monobath can be used for rapid in situ processing in a liquid gate for real-time holographic interferometry. This also permits utilization of a very simple solution handling system. Changes in emulsion thickness are reduced to an acceptable level and problems of matching refractive indices are eliminated by exposing and viewing the holograms in water. Excellent null patterns are obtained and real-time holographic interferometry can be carried out over long periods of time.