Identical Cores Research Articles

Optimization of throughput, fairness and energy efficiency on asymmetric multicore systems via OS scheduling.

Most of chip multiprocessors (CMPs) are symmetric, i.e. they are composed of identical cores. These CMPs may consist of complex cores (e.g., Intel Haswell or IBM Power8) or simple and lower-power cores (e.g. ARM Cortex A9 or Intel Xeon Phi). Cores in the former approach have advanced microarchitectural features, such as out-of-order super-scalar pipelines, and they are suitable for running sequential applications which use them efficiently. Cores in the latter approach have a simple microarchitecture and are good for running applications with high thread-level parallelism (TLP).

Design of a new low-phase-noise millimetre-wave quadrature voltage-controlled oscillator

ABSTRACTThis paper presents a new circuit topology of millimetre-wave quadrature voltage-controlled oscillator (QVCO) using an improved Colpitts oscillator without tail bias. By employing an extra capacitance between the drain and source terminations of the transistors and optimising circuit values, a low-power and low-phase-noise (PN) oscillator is designed. For generating the output signals with 90° phase difference, a self-injection coupling network between two identical cores is used. The proposed QVCO dissipates no extra dc power for coupling, since there is no dc-path to ground for the coupled transistors and no extra noise is added to circuit. The best figure-of-merit is −188.5, the power consumption is 14.98–15.45 mW, in a standard 180-nm CMOS technology, for 58.2 GHz center frequency from 59.3 to 59.6 GHz. The PN is −104.86 dBc/Hz at 1-MHz offset.

Modulation instability in collinear three-core optical fibers

Modulation instability (MI) in a collinear three-core optical fiber, namely, an optical fiber that consists of three identical cores aligned symmetrically along a straight line, is analyzed in detail for the symmetric and antisymmetric continuous-wave (CW) states in the normal and anomalous dispersion regimes. The symmetric CW state, where the fields in the two side cores are identical, exists only with certain combinations of powers in the side cores and the center core within a limited range of total power, and these characteristics are governed by the linear coupling and the nonlinear coefficient of the fiber. The symmetric CW state generates MI characteristics not observed previously. In the anomalous dispersion regime, as the total power increases from a threshold value, three MI bands are generated and then merge rapidly into a single band, and the MI gain is insensitive to the linear coupling coefficient and the coupling-coefficient dispersion (CCD) of the fiber. In the normal dispersion regime, the generation of MI depends critically on the CCD. Two MI bands are generated from the CCD at a total power smaller than a critical value that depends on the linear coupling, the dispersion, and the nonlinear coefficient of the fiber. No MI can be generated without CCD. The antisymmetric CW state, where the fields in the two side cores have equal amplitudes of opposite signs and the field in the center core is zero, exists at any power. For this state, the MI characteristics are qualitatively similar to those of an equilateral three-core fiber. The study also includes direct numerical verification of the MI analysis with wave propagation simulation.

Dual emission and optical gain in PbS/CdS nanocrystals: Role of shell volume and of core/shell interface

Colloidal semiconductor nanocrystals (NCs) emitting simultaneously at two distinct wavelengths are of great interest for multiple applications ranging from ultrasensitive self-calibrating nanoscale sensing to light-emitting diodes and white-light emission. The physical mechanism governing dual emission in core/shell NCs composed of two materials is still under investigation; in particular, the roles of the volume of the shell and of the core/shell interface are not well understood. Here we compare three different PbS/CdS NCs with identical cores, but shells of different thicknesses and different core/shell interfaces. We study photoluminescence (PL) characteristics and ultrafast exciton dynamics with subpicosecond time resolution. We find that dual emission is possible for both thick and thin quantum-confined shells, and for different core/shell interfaces. We show that very thick shells allow the decoupling of shell excitons from core ones and we fully describe the dynamic exciton barrier, connected to efficient shell PL. We discuss the efficiency of shell PL in relation to the properties of the NCs and show that it can give rise to optical gain. Our results provide a comprehensive understanding of the physical phenomena governing dual-emission mechanisms in NCs.

Neuromorphic Hardware Architecture Using the Neural Engineering Framework for Pattern Recognition.

We present a hardware architecture that uses the neural engineering framework (NEF) to implement large-scale neural networks on field programmable gate arrays (FPGAs) for performing massively parallel real-time pattern recognition. NEF is a framework that is capable of synthesising large-scale cognitive systems from subnetworks and we have previously presented an FPGA implementation of the NEF that successfully performs nonlinear mathematical computations. That work was developed based on a compact digital neural core, which consists of 64 neurons that are instantiated by a single physical neuron using a time-multiplexing approach. We have now scaled this approach up to build a pattern recognition system by combining identical neural cores together. As a proof of concept, we have developed a handwritten digit recognition system using the MNIST database and achieved a recognition rate of 96.55%. The system is implemented on a state-of-the-art FPGA and can process 5.12 million digits per second. The architecture and hardware optimisations presented offer high-speed and resource-efficient means for performing high-speed, neuromorphic, and massively parallel pattern recognition and classification tasks.

Two mm-Wave VCOs in 28-nm UTBB FD-SOI CMOS

Two 60-GHz band voltage controlled oscillators (VCOs) designed in a 28-nm ultrathin body and buried oxide fully depleted silicon on insulator (UTBB FD-SOI) CMOS process are demonstrated and compared. Both VCOs have identical cross-coupled nMOS cores and dissipate 3.15 mW from a 0.9-V supply. The first design uses a standard FET current source and achieves a figure of merit (FOM) of −181 dBc/Hz, whereas the second employs a filtered current source and achieves a state-of-the-art FOM of −187 dBc/Hz. The achieved 6-dB improvement demonstrates the efficiency of the filtering technique at millimeter wave frequencies and the feasibility of efficient low-phase noise designs in 28-nm UTBB FD-SOI CMOS. The active area of the filtered VCO is $90\,\,\mu \text{m}\,\,\times 180\,\,\mu \text{m}$ and the standard VCO has an area of $80\,\,\mu \text{m}\,\,\times 110\,\,\mu \text{m}$ .

SSAGA

The emergence of GPGPU applications, bolstered by flexible GPU programming platforms, has created a tremendous challenge in maintaining high energy efficiency in modern GPUs. In this article, we demonstrate that customizing a Streaming Multiprocessor (SM) of a GPU at a lower frequency is significantly more energy efficient compared to employing DVFS on an SM designed for a high-frequency operation. Using a system-level CAD technique, we propose SSAGA—Streaming Multiprocessors Synthesized for Asymmetric GPGPU Applications —an energy-efficient GPU design paradigm. SSAGA creates architecturally identical SM cores, customized for different voltage-frequency domains. Our rigorous cross-layer methodology demonstrates an average of 20% improvement in energy efficiency over a spatially multitasking GPU across a range of GPGPU applications.

Measuring the difference in group delays between the cores of a multicore optical fiber by means of time-domain interferometry

A way of measuring the difference in group delays between the cores of multi-core optical fiber is proposed. The method is based on a low-coherence Mach–Zehnder interferometer and a broadband radiation source. The corresponding experimental setup is constructed and a seven-core optical fiber with identical cores manufactured at the Fiber Optics Research Center, Russian Academy of Sciences, is studied.

Six-Mode 19-Core Fiber With 114 Spatial Modes for Weakly-Coupled Mode-Division-Multiplexed Transmission

This paper describes a six-mode 19-core fiber with 114 spatial modes, which was designed and fabricated for weakly-coupled mode-division-multiplexed (MDM) transmission over uncoupled cores. The 19 identical graded-index cores were packed on a hexagonal lattice with a 62-μm pitch in a ∼318-μm-diameter cladding. The cladding diameter is large, but still mechanically reliable by assuming a 2% proof test and 40-mm minimum fiber bend radius. The inter-core crosstalk is negligibly suppressed and the calculation indicated that a 50-μm core pitch (corresponding to the 266-μm-diameter cladding) is achievable without significant crosstalk degradation. Though the fiber was intended for the weakly-coupled MDM transmission, the differential mode delay (DMD) within a degenerated mode group of LP21 and LP02 modes was not well suppressed. Thus, further DMD suppression within each mode group or splitting the degeneracy of the modes is required to realize multiple-input multiple-output (MIMO) digital signal processing with reduced complexity. Additionally, a numerical investigation revealed that the intra-core inter-modal crosstalk at a butt coupling was induced not only by the misalignment, but also by the refractive index profile mismatch between the butt-coupled cores. Therefore, the tight tolerance on the refractive index profile is important both for DMD control and modal crosstalk suppression.

Core/Wavelength Selective Switching Based on Heterogeneous MCFs With LPGs

In this letter, the light transfer promoted by long-period gratings (LPGs) inscribed in heterogeneous multicore fibers (HeMCFs) is theoretically analyzed. We determined the power evolution along the LPG in two different four-core HeMCFs. This letter demonstrates the selective switch of optical power between the identical cores of the HeMCFs with less than −40 dB of crosstalk to the other cores, regardless the position of the identical cores. The results show that the proposed technique can be used to produce inline selective core switching for multicore space division multiplexing.

Parallelized Network-on-Chip-Reused Test Access Mechanism for Multiple Identical Cores

This paper proposes a new network-on-chip (NoC)-reused test access mechanism (TAM) for testing multiple identical cores. It can test multiple cores concurrently and identify faulty cores to derate the chip by excluding the core. In order to minimize the test time, the TAM utilizes the majority value of test response data. All of the cores can thereby be tested in parallel and test costs (in both test pins and test time) are exactly the same as those for a single core. The hardware overhead is minimized by reusing the NoC infrastructures and transfer-counters are designed as a majority analyzer. The experimental results in this paper show that the proposed TAM can test multiple cores in the same time as a single core and with negligible hardware overhead.

Electrical characterization of individual metal‐coated polymer spheres used in isotropic conductive adhesives

ABSTRACTIsotropic conductive adhesives (ICAs) filled with metal‐coated polymer spheres (MPS) have been proposed to improve the mechanical reliability compared to conventional ICAs filled with silver flakes. The electrical properties of MPS play an important role in the electrical performance of macroscopic MPS‐based ICAs. This article deals with the electrical characterization of individual MPS using a nanoindentation‐based flat punch method, in which the resistance and the deformation of single MPS were monitored simultaneously. Four groups of silver‐coated polymer spheres (AgPS) with identical polymer cores but different silver coating thicknesses were tested. The resistance of AgPS decreases gradually with increasing deformation degree of particles, and increases when the deformation of particles is reduced. In addition, the resistance of individual AgPS is dependent on the physical properties of the silver coating, such as thickness, uniformity, and porosity. The thicker the silver coating is, the lower and more stable the resistance of AgPS is. © 2016 Wiley Periodicals, Inc. J. Appl. Polym. Sci. 2016, 133, 43764.

Influence of fibre structure and bends on optical cross-talk in multicore fibres

We have studied the influence of cross-sectional structure and bends on optical cross-talk in a multicore fibre. A reduced refractive index layer produced between the cores of such fibre with a small centre-to-centre spacing between neighbouring cores reduces optical cross-talk by . The cross-talk level achieved, per kilometre of the length of the multicore fibre, is acceptable for a number of applications where relatively small lengths of fibre are needed. Moreover, a significant decrease in optical cross-talk has been ensured by reducing the winding diameter of multicore fibres with identical cores.

Approximation algorithms for scheduling on multi-core processor with shared speedup resources

We consider a joint resource partition and scheduling problem. We are given m identical cores and discrete resources of total size k. We need to partition the resources among these cores. A set of jobs must be processed non-preemptively on these cores after the resource partition. The processing time of a job on a core depends on the size of resources allocated to that corresponding core. The resource allocation scheme is static, i.e., we cannot change the amount of resources that was allocated to a core during the whole scheduling. Hassidim et al. (2013) investigated this problem with a general processing time function, i.e., the processing time of a job is an arbitrary function of the level of resources allocated to that core. They provided an algorithm with approximation ratio of 36. In this paper, we improve the approximation ratio to 8 by presenting a new resource partition scheme. Next, we consider a special model where the core’s speed is proportional to its allocated resource, then we present two algorithms with improved approximation ratios.

Chemical composition of hybrid silicon nanoparticles and ultrafast dynamics of charge carriers

The qualitative and quantitative composition of hybrid Si/SiO x /SiO2/OH(D) nanoparticles synthesized from silicon monoxide has been determined by X-ray photoelectron spectroscopy. The nanoparticles are composed of a Si crystalline core and a SiO x /SiO2/OH(D) shell, where SiO x is the interface of intermediate oxides corresponding to the Si1+, Si2+, and Si3+ valence states of silicon and SiO2/OH(D) is the external shell of the nanoparticle. The chemical composition and average stoichiometry of deuterated and nondeuterated nanoparticles have been determined; the identified compositions have been compared. The dependence of photoluminescent properties on the composition of the samples has been discussed. Two forms of hydrophilic silicon nanoparticles with identical crystalline cores—photoluminescent deuteriumpassivated particles oxidized in completely deuterated dimethylsulfoxide and nonphotoluminescent hydrogen- passivated reference samples oxidized in dimethylsulfoxide—have been studied by broadband femtosecond spectroscopy. It has been found that there are significant differences in the ultrafast spectral–temporal induced absorption dynamics of these two forms in the energy range corresponding to the calculated bandgap of the nanoparticles. The observed difference has been attributed to the specific features of the relaxation of excited charge carriers in the energy states responsible for photoluminescence in the red spectral region.

Compact quantum dot–antibody conjugates for FRET immunoassays with subnanomolar detection limits

A novel two-step approach for quantum dot (QD) functionalization and bioconjugation is presented, which yields ultra-compact, stable, and highly luminescent antibody-QD conjugates suitable for use in FRET immunoassays. Hydrophobic InPZnS/ZnSe/ZnS (emission wavelength: 530 nm), CdSe/ZnS (605 nm), and CdSeTe/ZnS (705 nm) QDs were surface functionalized with zwitterionic penicillamine, enabling aqueous phase transfer under conservation of the photoluminescence properties. Post-functionalization with a heterobifunctional crosslinker, containing a lipoic acid group and a maleimide function, enabled the subsequent coupling to sulfhydryl groups of proteins. This was demonstrated by QD conjugation with fragmented antibodies (F(ab)). The obtained F(ab)-QD conjugates range among the smallest antibody-functionalized nanoprobes ever reported, with a hydrodynamic diameter <13 nm, PL quantum yield up to 66% at 705 nm, and colloidal stability of several months in various buffers. They were applied as FRET acceptors in homogeneous, time-gated immunoassays using Tb-antibodies as FRET donors, both coupled by an immunological sandwich complex between the two antibodies and a PSA (prostate specific antigen) biomarker. The advantages of the compact surface coating for FRET could be demonstrated by an 6.2 and 2.5 fold improvement of the limit of detection (LOD) for PSA compared to commercially available hydrophilic QDs emitting at 605 and 705 nm, respectively. While the commercial QDs contain identical inorganic cores responsible for their fluorescence, they are coated with a comparably thick amphiphilic polymer layer leading to much larger hydrodynamic diameters (>26 nm without biomolecules). The LODs of 0.8 and 3.7 ng mL(-1) obtained in 50 μL serum samples are below the clinical cut-off level of PSA (4 ng mL(-1)) and demonstrate their direct applicability in clinical diagnostics.

A Common Mode Inductor With External Magnetic Field Immunity, Low-Magnetic Field Emission, and High-Differential Mode Inductance

This paper proposes a stacked common-mode (CM) inductor for electromagnetic interference (EMI) attenuation. The proposed stacked CM inductor consists of two CM inductors with identical cores and the same number of winding turns but with opposite winding directions. The proposed stacked CM inductor has three advantages. First, it is much less susceptible to external magnetic field interference than conventional CM inductors. This suppresses the EMI due to the near-magnetic-field coupling to the CM inductor. Second, the proposed winding structure effectively increases the leakage inductance of the CM inductor, which enhances its ability to suppress differential-mode (DM) noise. Third, the magnetic field generated by the DM current in the stacked CM inductor is partially canceled outside of the inductor. As a result, the stacked CM inductor has lower magnetic field emission than that of a conventional CM inductor. Simulation and experimental results validate the advantages of the proposed CM inductor, which show the proposed technique greatly improves the performance of CM inductors.

Chromatic Dispersion of an Optical Fiber Based on Photonic Quasicrystals with Twelve-Fold Symmetry and its Application as Directional Coupling

Optical fibers composed by photonic quasicrystals it is based on aperiodic structures characterized by at least two different symmetrical patterns from the base matrix. In this work, an optical fiber composed by quasi-periodic and symmetric matrix was analyzed using the finite element method in conjunction with cylindrical perfectly matched layers. The structure is composed by pure silica and/or silica doped with germanium, and originated from twelve distributions of air holes symmetrically with a defect caused by the absence of the fiber central hole. The obtained results shown that the structure in analysis exhibits an ultra-flat chromatic dispersion for a range of wavelengths varying from 1.4 to 1.6 µm covering the bands E,S, C and L, with chromatic dispersion, for the silica doped with germanium, varying between 22.1 and 23.01 [ps.km-1.nm-1]. This photonic quasicrystal fiber (PQCF) was used to form a coupler with three identical horizontal cores surrounded by air holes. The signal power launched in the central core is equally divided between the two neighboring cores with 50% of the coupling ratio.

The Effect of Common-Mode Voltage Elimination on the Iron Loss in Machine Core Laminations of Multilevel Drives

This paper studies the effect of common-mode voltage elimination (CMVE) on the iron loss of electrical machine core laminations under multilevel converter supply. Three identical magnetic ring cores are excited by either a three-level converter or a five-level voltage source converter to study the behavior of CMVE on a three-phase system. Both multilevel converters are controlled by using a space vector pulsewidth modulation as it is one of the most often used techniques for CMVE. These experimental results are confirmed numerically with a dynamic iron loss model. In addition, the effect of CMVE, at different switching frequencies, on the core loss of a synchronous machine is numerically studied. The results presented in this paper show that the core loss is considerably increased when the CMVE is implemented. However, this iron loss increase in the five-level drive systems is lower compared with the three-level ones. Therefore, it is important that the designers of drive systems take such effects into consideration.

Formation of accretion centers in simulations of colliding uniform density H2 cores

AbstractWe test here the first stage of a route of modifications to be applied to the public GADGET2 code for dynamically identifying accretion centers during the collision process of two adjacent and identical gas cores. Each colliding core has a uniform density profile and rigid body rotation; its mass and size have been chosen to represent the observed core L1544; for the thermal and rotational energy ratios with respect to the potential energy, we assume the values α = 0.3 and β = 0.1, respectively. These values favor the gravitational collapse of the core. We here study cases of both head‐on and off‐center collisions, in which the pre‐collision velocity increases the initial sound speed of the barotropic gas by up to several times. In a simulation the accretion centers are formed by the highest density particles, so we here report their location and properties in order to realize the collision effects on the collapsing and colliding cores. In one of the models, we observe a roughly spherical distribution of accretion centers located at the front wave of the collision. In a forthcoming publication we will apply the full modified GADGET code to study the collision of turbulent cores. (© 2015 WILEY‐VCH Verlag GmbH &amp; Co. KGaA, Weinheim)