Search

Commodity multicore systems are increasingly adopting hardware support that enables the system software to partition the last-level cache (LLC). This support makes it possible for the operating system (OS) or the Virtual Machine Monitor (VMM) to mitigate shared-resource contention effects on multicores by assigning different co-running applications to various cache partitions. Recently cache-clustering (or partition-sharing) strategies have emerged as a way to improve system throughput and fairness on new platforms with cache-partitioning support. As opposed to strict cache-partitioning, which allocates separate cache partitions to each application, cache-clustering allows partitions to be shared by a group of applications. In this article we propose LFOC+, a fairness-aware OS-level cache-clustering policy for commodity multicore systems. LFOC+ tries to mimic the behavior of the optimal cache-clustering solution for fairness, which we could obtain for different workload scenarios by using a simulation tool. Our dynamic cache-clustering strategy continuously gathers data from performance monitoring counters to classify applications at runtime based on the degree of cache sensitivity and contentiousness, and effectively separates cache-sensitive applications from aggressor programs to improve fairness, while providing acceptable system throughput. We implemented LFOC+ in the Linux kernel and evaluated it on a real system featuring an Intel Skylake processor, where we compare its effectiveness to that of four previously proposed cache-clustering policies. Our experimental analysis reveals that LFOC+ constitutes a lightweight OS-level policy and improves fairness relative to two other state-of-the-art fairness-aware strategies --Dunn and LFOC--, by up to 22\% and up to 20.6\%, respectively, and by 9\% and 4.9\% on average.

Asymmetric multicore processors (AMPs) couple high-performance big cores and low-power small cores with the same instruction-set architecture but different features, such as clock frequency or microarchitecture. Previous work has shown that asymmetric designs may deliver higher energy efficiency than symmetric multicores for diverse workloads. Despite their benefits, AMPs pose significant challenges to runtime systems of parallel programming models. While previous work has mainly explored how to efficiently execute task-based parallel applications on AMPs, via enhancements in the runtime system, improving the performance of unmodified data-parallel applications on these architectures is still a big challenge. In this work we analyze the particular case of loop-based OpenMP applications, which are widely used today in scientific and engineering domains, and constitute the dominant application type in many parallel benchmark suites used for performance evaluation on multicore systems. We observed that conventional loop-scheduling OpenMP approaches are unable to efficiently cope with the load imbalance that naturally stems from the different performance delivered by big and small cores. To address this shortcoming, we propose \textit{Asymmetric Iteration Distribution} (AID), a set of novel loop-scheduling methods for AMPs that distribute iterations unevenly across worker threads to efficiently deal with performance asymmetry. We implemented AID in \textit{libgomp} --the GNU OpenMP runtime system--, and evaluated it on two different asymmetric multicore platforms. Our analysis reveals that the AID methods constitute effective replacements of the \texttt{static} and \texttt{dynamic} methods on AMPs, and are capable of improving performance over these conventional strategies by up to 56\% and 16.8\%, respectively.

Multicore processors constitute the main architecture choice for modern computing systems in different market segments. Despite their benefits, the contention that naturally appears when multiple applications compete for the use of shared resources among cores, such as the last-level cache (LLC), may lead to substantial performance degradation. This may have a negative impact on key system aspects such as throughput and fairness. Assigning the various applications in the workload to separate LLC partitions with possibly different sizes, has been proven effective to mitigate shared-resource contention effects. In this article we propose LFOC, a clustering-based cache partitioning scheme that strives to deliver fairness while providing acceptable system throughput. LFOC leverages the Intel Cache Allocation Technology (CAT), which enables the system software to divide the LLC into different partitions. To accomplish its goals, LFOC tries to mimic the behavior of the optimal cache-clustering solution, which we could approximate by means of a simulator in different scenarios. To this end, LFOC effectively identifies streaming aggressor programs and cache sensitive applications, which are then assigned to separate cache partitions. We implemented LFOC in the Linux kernel and evaluated it on a real system featuring an Intel Skylake processor, where we compare its effectiveness to that of two state-of-the-art policies that optimize fairness and throughput, respectively. Our experimental analysis reveals that LFOC is able to bring a higher reduction in unfairness by leveraging a lightweight algorithm suitable for adoption in a real OS.

Various constraints of Static Random Access Memory (SRAM) are leading to consider new memory technologies as candidates for building on-chip shared last-level caches (SLLCs). Spin-Transfer Torque RAM (STT-RAM) is currently postulated as the prime contender due to its better energy efficiency, smaller die footprint and higher scalability. However, STT-RAM also exhibits some drawbacks, like slow and energy-hungry write operations, that need to be mitigated. In this work we address these shortcomings by leveraging a new management mechanism for STT-RAM SLLCs. This approach is based on the previous observation that the stream of references arriving at the SLLC of a Chip MultiProcessor (CMP) exhibits reuse locality, i.e., those blocks referenced several times manifest high probability of forthcoming reuse. In this paper, we employ a cache management mechanism that selects the contents of the SLLC aimed to exploit reuse locality instead of temporal locality. Specifically, our proposal consists in the inclusion of a Reuse Detector between private cache levels and the STT-RAM SLLC to detect blocks that do not exhibit reuse, in order to avoid their insertion in the SLLC, hence reducing the number of write operations and the energy consumption in the STT-RAM. Our evaluation reveals that our scheme reports on average, energy reductions in the SLLC in the range of 37-30\%, additional energy savings in the main memory in the range of 6-8\% and performance improvements of 3\% up to 14\% (16-core) compared to an STT-RAM SLLC baseline where no reuse detector is employed. More importantly, our approach outperforms DASCA, the state-of-the-art STT-RAM SLLC management, reporting SLLC energy savings in the range of 4-11\% higher than those of DASCA, delivering higher performance in the range of 1.5-14\%, and additional improvements in DRAM energy consumption in the range of 2-9\% higher than DASCA.

The heterogeneous computing paradigm has led to the need for portable and efficient programming solutions that can leverage the capabilities of various hardware devices, such as NVIDIA, Intel, and AMD GPUs. This study evaluates the portability and performance of the SYCL and CUDA languages for one fundamental bioinformatics application (Smith-Waterman protein database search) across different GPU architectures, considering single and multi-GPU configurations from different vendors. The experimental work showed that, while both CUDA and SYCL versions achieve similar performance on NVIDIA devices, the latter demonstrated remarkable code portability to other GPU architectures, such as AMD and Intel. Furthermore, the architectural efficiency rates achieved on these devices were superior in 3 of the 4 cases tested. This brief study highlights the potential of SYCL as a viable solution for achieving both performance and portability in the heterogeneous computing ecosystem., Comment: This article was accepted for publication in 2023 IEEE 35th International Symposium on Computer Architecture and High Performance Computing (SBAC-PAD)

The accuracy requirements in many scientific computing workloads result in the use of double-precision floating-point arithmetic in the execution kernels. Nevertheless, emerging real-number representations, such as posit arithmetic, show promise in delivering even higher accuracy in such computations. In this work, we explore the native use of 64-bit posits in a series of numerical benchmarks and compare their timing performance, accuracy and hardware cost to IEEE 754 doubles. In addition, we also study the conjugate gradient method for numerically solving systems of linear equations in real-world applications. For this, we extend the PERCIVAL RISC-V core and the Xposit custom RISC-V extension with posit64 and quire operations. Results show that posit64 can obtain up to 4 orders of magnitude lower mean square error than doubles. This leads to a reduction in the number of iterations required for convergence in some iterative solvers. However, leveraging the quire accumulator register can limit the order of some operations such as matrix multiplications. Furthermore, detailed FPGA and ASIC synthesis results highlight the significant hardware cost of 64-bit posit arithmetic and quire. Despite this, the large accuracy improvements achieved with the same memory bandwidth suggest that posit arithmetic may provide a potential alternative representation for scientific computing., Comment: 12 pages. Code available at https://github.com/artecs-group/PERCIVAL. Reviewed version to be published in IEEE TC

The bacterium Xylella fastidiosa (Xf) is mainly transmitted by the spittlebug, Philaenus spumarius, in Europe, where it has caused significant economic damage to olive and almond trees. Understanding the factors that determine disease dynamics in pathosystems that share similarities can help design control strategies focused on minimizing transmission chains. Here we introduce a compartmental model for Xf-caused diseases in Europe that accounts for the main relevant epidemiological processes, including the seasonal dynamics of P. spumarius. The model was confronted with epidemiological data from the two major outbreaks of Xf in Europe, the olive quick disease syndrome (OQDS) in Apulia, Italy, caused by the subspecies pauca, and the almond leaf scorch disease (ALSD) in Majorca, Spain, caused by subspecies multiplex and fastidiosa. Using a Bayesian inference framework, we show how the model successfully reproduces the general field data in both diseases. In a global sensitivity analysis, the vector-plant and plant-vector transmission rates, together with the vector removal rate, were the most influential parameters in determining the time of the infected host population peak, the incidence peak and the final number of dead hosts. We also used our model to check different vector-based control strategies, showing that a joint strategy focused on increasing the rate of vector removal while lowering the number of annual newborn vectors is optimal for disease control., Comment: 18 pages, 8 figures

Bioinformatics and Computational Biology are two fields that have been exploiting GPUs for more than two decades, being CUDA the most used programming language for them. However, as CUDA is an NVIDIA proprietary language, it implies a strong portability restriction to a wide range of heterogeneous architectures, like AMD or Intel GPUs. To face this issue, the Khronos Group has recently proposed the SYCL standard, which is an open, royalty-free, cross-platform abstraction layer, that enables the programming of a heterogeneous system to be written using standard, single-source C++ code. Over the past few years, several implementations of this SYCL standard have emerged, being oneAPI the one from Intel. This paper presents the migration process of the SW\# suite, a biological sequence alignment tool developed in CUDA, to SYCL using Intel's oneAPI ecosystem. The experimental results show that SW\# was completely migrated with a small programmer intervention in terms of hand-coding. In addition, it was possible to port the migrated code between different architectures (considering multiple vendor GPUs and also CPUs), with no noticeable performance degradation on 5 different NVIDIA GPUs. Moreover, performance remained stable when switching to another SYCL implementation. As a consequence, SYCL and its implementations can offer attractive opportunities for the Bioinformatics community, especially considering the vast existence of CUDA-based legacy codes.

Emerging marine infectious diseases pose a substantial threat to marine ecosystems and the conservation of their biodiversity. Compartmental models of epidemic transmission in marine sessile organisms, available only recently, are based on non-spatial descriptions in which space is homogenised and parasite mobility is not explicitly accounted for. However, in realistic scenarios epidemic transmission is conditioned by the spatial distribution of hosts and the parasites mobility patterns, calling for a explicit description of space. In this work we develop a spatially-explicit individual-based model to study disease transmission by waterborne parasites in sessile marine populations. We investigate the impact of spatial disease transmission through extensive numerical simulations and theoretical analysis. Specifically, the effects of parasite mobility into the epidemic threshold and the temporal progression of the epidemic are assessed. We show that larger values of pathogen mobility imply more severe epidemics, as the number of infections increases, and shorter time-scales to extinction. An analytical expression for the basic reproduction number of the spatial model is derived as function of the non-spatial counterpart, which characterises a transition between a disease-free and a propagation phase, in which the disease propagates over a large fraction of the system., Comment: 11 pages, 6 figures, 1 table

To face the programming challenges related to heterogeneous computing, Intel recently introduced oneAPI, a new programming environment that allows code developed in Data Parallel C++ (DPC++) language to be run on different devices such as CPUs, GPUs, FPGAs, among others. To tackle CUDA-based legacy codes, oneAPI provides a compatibility tool (dpct) that facilitates the migration to DPC++. Due to the large amount of existing CUDA-based software in the bioinformatics context, this paper presents our experiences porting SW#db, a well-known sequence alignment tool, to DPC++ using dpct. From the experimental work, it was possible to prove the usefulness of dpct for SW#db code migration and the cross-GPU vendor, cross-architecture portability of the migrated DPC++ code. In addition, the performance results showed that the migrated DPC++ code reports similar efficiency rates to its CUDA-native counterpart or even better in some tests (approximately +5%)., Comment: Accepted in IWBBIO 2022

We have studied the existence of traveling pulses in a general Type-I excitable 1-dimensional medium. We have obtained the stability region and characterized the different bifurcations behind either the destruction or loss of stability of the pulses. In particular, some of the bifurcations delimiting the stability region have been connected, using singular limits, with the two different scenarios that mediated the Type-I local excitability, i.e. homoclinic (saddle-loop) and Saddle-Node on the Invariant Circle bifurcations. The existence of the traveling pulses has been linked, outside the stability region, to a drift pitchfork instability of localized steady structures., Comment: 17 pages, 12 figures

Since the last century, deterministic compartmental models have emerged as powerful tools to predict and control epidemic outbreaks, in many cases helping to mitigate their impacts. A key quantity for these models is the so-called Basic Reproduction Number, that measures the number of secondary infections produced by an initial infected individual in a fully susceptible population. Some methods have been developed to allow the direct computation of this quantity provided that some conditions are fulfilled, such that the model has a pre-pandemic disease-free equilibrium state. This condition is only fulfilled when the populations are stationary. In the case of vector-borne diseases, this implies that the vector birth and death rates need to be balanced, what is not fulfilled in many realistic cases in which the vector population grow or decrease. Here we develop a vector-borne epidemic model with growing and decaying vector populations and study the conditions under which the standard methods to compute $R_0$ work and discuss an alternative when they fail. We also show that growing vector populations produce a delay in the epidemic dynamics when compared to the case of the stationary vector population. Finally, we discuss the conditions under which the model can be reduced to the SIR model with fewer compartments and parameters, which helps in solving the problem of parameter unidentifiability of many vector-borne epidemic models., Comment: 16 pages, 6 figures

The state of the art of epidemic modelling in terrestrial ecosystems is the compartmental SIR model and its extensions from the now classical work of Kermack-Mackendrick. In contrast, epidemic modelling of marine ecosystems is a bit behind, and compartmental models have been introduced only recently. One of the reasons is that many epidemic processes in terrestrial ecosystems can be described through a contact process, while modelling marine epidemics is more subtle in many cases. Here we present a model describing disease outbreaks caused by parasites in bivalve populations. The SIRP model is a multicompartmental model with four compartments, three of which describe the different states of the host, susceptible (i.e. healthy), S, infected, I, and removed (dead), R, and one compartment for the parasite in the marine medium, P, written as a 4-dimensional dynamical system. Even if this is the simplest model one can write to describe this system, it is still too complicated for both direct analytical manipulation and direct comparison with experimental observations, as it depends on four parameters to be fitted. We show that it is possible to simplify the model, including a reduction to the standard SIR model if the parameters fulfil certain conditions. The model is validated with available data for the recent Mass Mortality Event of the noble pen shell Pinna nobilis, a disease caused by the parasite Haplosporidium pinnae, showing that the reduced SIR model is able to fit the data. So, we show that a model in which the species that suffers the epidemics (host) cannot move, and contagion occurs through parasites, can be reduced to the standard SIR model that represents epidemic transmission between mobile hosts. The fit indicates that the assumptions made to simplify the model are reasonable in practice, although it leads to an indeterminacy in three of the original parameters., Comment: 26 pages, 9 figures

We study Class-I excitable $1$-dimensional media showing the appearance of propagating traveling pulses. We consider a general model exhibiting Class-I excitability mediated by two different scenarios: a homoclinic (saddle-loop) and a SNIC (Saddle-Node on the Invariant Circle) bifurcations. The distinct properties of Class-I with respect to Class-II excitability infer unique properties to traveling pulses in Class-I excitable media. We show how the pulse shape inherit the infinite period of the homoclinic and SNIC bifurcations at threshold, exhibiting scaling behaviors in the spatial thickness of the pulses that are equivalent to the scaling behaviors of characteristic times in the temporal case., Comment: 5 pages, 6 figures

Automatic digitization of chess games using computer vision is a significant technological challenge. This problem is of much interest for tournament organizers and amateur or professional players to broadcast their over-the-board (OTB) games online or analyze them using chess engines. Previous work has shown promising results, but the recognition accuracy and the latency of state-of-the-art techniques still need further enhancements to allow their practical and affordable deployment. We have investigated how to implement them on an Nvidia Jetson Nano single-board computer effectively. Our first contribution has been accelerating the chessboard's detection algorithm. Subsequently, we have analyzed different Convolutional Neural Networks for chess piece classification and how to map them efficiently on our embedded platform. Notably, we have implemented a functional framework that automatically digitizes a chess position from an image in less than 1 second, with 92% accuracy when classifying the pieces and 95% when detecting the board., Comment: The complete source code of the LiveChess2FEN framework is publicly available with an open-source license in our GitHub repository: https://github.com/davidmallasen/LiveChess2FEN

The well-known Smith-Waterman (SW) algorithm is the most commonly used method for local sequence alignments. However, SW is very computationally demanding for large protein databases. There exist several implementations that take advantage of computing parallelization on many-cores, FPGAs or GPUs, in order to increase the alignment throughtput. In this paper, we have explored SW acceleration on Intel KNL processor. The novelty of this architecture requires the revision of previous programming and optimization techniques on many-core architectures. To the best of authors knowledge, this is the first KNL architecture assessment for SW algorithm. Our evaluation, using the renowned Environmental NR database as benchmark, has shown that multi-threading and SIMD exploitation reports competitive performance (351 GCUPS) in comparison with other implementations., Comment: Submitted to Euro-Par 2017 Conference

In this paper we describe HeSP, a complete simulation framework to study a general task scheduling-partitioning problem on heterogeneous architectures, which treats recursive task partitioning and scheduling decisions on equal footing. Considering recursive partitioning as an additional degree of freedom, tasks can be dynamically partitioned or merged at runtime for each available processor type, exposing additional or reduced degrees of parallelism as needed. Our simulations reveal that, for a specific class of dense linear algebra algorithms taken as a driving example, simultaneous decisions on task scheduling and partitioning yield significant performance gains on two different heterogeneous platforms: a highly heterogeneous CPU-GPU system and a low-power asymmetric big.LITTLE ARM platform. The insights extracted from the framework can be further applied to actual runtime task schedulers in order to improve performance on current or future architectures and for different task-parallel codes.

We study the influence of a linear nonlocal spatial coupling on the interaction of fronts connecting two equivalent stable states in the prototypical 1-D real Ginzburg-Landau equation. While for local coupling the fronts are always monotonic and therefore the dynamical behavior leads to coarsening and the annihilation of pairs of fronts, nonlocal terms can induce spatial oscillations in the front, allowing for the creation of localized structures, emerging from pinning between two fronts. We show this for three different nonlocal influence kernels. The first two, mod-exponential and Gaussian, are positive-definite and decay exponentially or faster, while the third one, a Mexican-hat kernel, is not positive definite., Comment: 16 pages, 14 figures

The present work studies the influence of nonlocal spatial coupling on the existence of localized structures in 1-dimensional extended systems. We consider systems described by a real field with a nonlocal coupling that has a linear dependence on the field. Leveraging spatial dynamics we provide a general framework to understand the effect of the nonlocality on the shape of the fronts connecting two stable states. In particular we show that non local terms can induce spatial oscillations in the front tails, allowing for the creation of localized structures, emerging from pinning between two fronts. In parameter space the region where fronts are oscillatory is limited by three transitions: the modulational instability of the homogeneous state, the Belyakov-Devaney transition in which monotonic fronts acquire spatial oscillations with infinite wavelength, and a crossover in which monotonically decaying fronts develop oscillations with a finite wavelength. We show how these transitions are organized by codimension 2 and 3 point s and illustrate how by changing the parameters of the nonlocal coupling it is possible to bring the system into the region where localized structures can be formed., Comment: 14 pages, 9 figures

We show how to exploit excitable regimes mediated by localized structures (LS) to perform AND, OR, and NOT logical operations providing full logical functionality. Our scheme is general and can be implemented in any physical system displaying LS. In particular, LS in nonlinear photonic devices can be used for all-optical computing applications where several reconfigurable logic gates can be implemented in the transverse plane of a single device, allowing for parallel computing., Comment: 11 pages, 6 figures

We demonstrate that nonlocal coupling strongly influences the dynamics of fronts connecting two equivalent states. In two prototype models we observe a large amplification in the interaction strength between two opposite fronts increasing front velocities several orders of magnitude. By analyzing the spatial dynamics we prove that way beyond quantitative effects, nonlocal terms can also change the overall qualitative picture by inducing oscillations in the front profile. This leads to a mechanism for the formation of localized structures not present for local interactions. Finally, nonlocal coupling can induce a steep broadening of localized structures, eventually annihilating them., Comment: 4 pages, 6 figures

We study and characterize a direct route to high-dimensional chaos (i.e. not implying an intermediate low-dimensional attractor) of a system composed out of three coupled Lorenz oscillators. A geometric analysis of this medium-dimensional dynamical system is carried out through a variety of numerical quantitative and qualitative techniques, that ultimately lead to the reconstruction of the route. The main finding is that the transition is organized by a heteroclinic explosion. The observed scenario resembles the classical route to chaos via homoclinic explosion of the Lorenz model., Comment: 12 pages, 11 figures, 1 table. Extended version of [D. Pazo & M.A. Matias, EPL 72, 176-182 (2005)]

In this work we characterize the dynamical instabilities of localized structures exhibited by a recently introduced Gelens et al., Phys. Rev. A 75, 063812 2007 generalization of the Lugiato-Lefever model that includes a weakly nonlocal response of an intracavity metamaterial. A rich scenario, in which the localized structures exhibit different types of oscillatory instabilities, tristability, and excitability, including a regime of conditional excitability in which the system is bistable, is presented and discussed. Finally, it is shown that the scenario is organized by a pair of Takens-Bogdanov codimension-2 points., Comment: 8 pages, 9 figures

We study the dynamical behavior of dissipative solitons in an optical cavity filled with a Kerr medium when a localized beam is applied on top of the homogeneous pumping. In particular, we report on the excitability regime that cavity solitons exhibits which is emergent property since the system is not locally excitable. The resulting scenario differs in an important way from the case of a purely homogeneous pump and now two different excitable regimes, both Class I, are shown. The whole scenario is presented and discussed, showing that it is organized by three codimension-2 points. Moreover, the localized beam can be used to control important features, such as the excitable threshold, improving the possibilities for the experimental observation of this phenomenon., Comment: 9 Pages, 12 figures

In this work we demonstrate for an experimental system, that exhibits the Lorenz butterfly attractor behavior, that perfect chaotic phase synchronization cannot be achieved in systems with an unbounded distribution of intrinsic time scales. Instead, imperfect phase synchronization is characterized by the occurrence of phase slips, associated to epochs of time during which the chaotic oscillator exhibits a slower time scale. Interestingly, during phase slips the chaotic oscillator keeps in sync with the drive, but with a different locking ratio., Comment: 7 pages, 8 figures

In this work we characterize in detail the bifurcation leading to an excitable regime mediated by localized structures in a dissipative nonlinear Kerr cavity with a homogeneous pump. Here we show how the route can be understood through a planar dynamical system in which a limit cycle becomes the homoclinic orbit of a saddle point (saddle-loop bifurcation). The whole picture is unveiled, and the mechanism by which this reduction occurs from the full infinite-dimensional dynamical system is studied. Finally, it is shown that the bifurcation leads to an excitability regime, under the application of suitable perturbations. Excitability is an emergent property for this system, as it emerges from the spatial dependence since the system does not exhibit any excitable behavior locally., Comment: 10 pages, 9 figures

At the eight-year anniversary of Watts & Strogatz's work on the collective dynamics of small-world networks and seven years after Barabasi & Albert's discovery of scale-free networks, the area of dynamical processes on complex networks is at the forefront of the current research on nonlinear dynamics and complex systems. This volume brings together a selection of original contributions in complementary topics of statistical physics, nonlinear dynamics and biological sciences, and is expected to provide the reader with a comprehensive up-to-date representation of this rapidly developing area., Comment: Preface article of the Physica D Special Issue "Dynamics on Complex Networks and Applications" (4 pages). Full issue available at http://www.sciencedirect.com/science/journal/01672789

Clonal organisms present a particular challenge in population genetics because, in addition to the possible existence of replicates of the same genotype in a given sample, some of the hypotheses and concepts underlying classical population genetics models are irreconcilable with clonality. The genetic structure and diversity of clonal populations was examined using a combination of new tools to analyze microsatellite data in the marine angiosperm Posidonia oceanica. These tools were based on examination of the frequency distribution of the genetic distance among ramets, termed the spectrum of genetic diversity (GDS), and of networks built on the basis of pairwise genetic distances among genets. The properties and topology of networks based on genetic distances showed a "small-world" topology, characterized by a high degree of connectivity among nodes, and a substantial amount of substructure, revealing organization in sub-families of closely related individuals. Keywords: genetic networks; small-world networks; genetic diversity; clonal organisms, Comment: Replaced with revised version

I show that the conclusions of [Hwang, Chavez, Amann, & Boccaletti, PRL 94, 138701 (2005); Chavez, Hwang, Amann, Hentschel, & Boccaletti, PRL 94, 218701 (2005)] are closely related to those of previous publications., Comment: 2 pages