Exploiting Programmable Network Interfaces for Parallel Query Execution in Workstation Clusters

Workstation clusters equipped with high performance interconnect having programmable network processors facilitate interesting opportunities to enhance the performance of parallel application run on them. In this paper, we propose schemes where certain application level processing in parallel database query execution is performed on the network processor. We evaluate the performance of TPC-H queries executing on a high end cluster where all tuple processing is done on the host processor, using a timed Petri net model, and find that tuple processing costs on the host processor dominate the execution time. These results are validated using a small cluster. We therefore propose 4 schemes where certain tuple processing activity is offloaded to the network processor. The first 2 schemes offload the tuple splitting activity - computation to identify the node on which to process the tuples, resulting in an execution time speedup of 1.09 relative to the base scheme, but with I/O bus becoming the bottleneck resource. In the 3rd scheme in addition to offloading tuple processing activity, the disk and network interface are combined to avoid the I/O bus bottleneck, which results in speedups up to 1.16, but with high host processor utilization. Our 4th scheme where the network processor also performs apart of join operation along with the host processor, gives a speedup of 1.47 along with balanced system resource utilizations. Further we observe that the proposed schemes perform equally well even in a scaled architecture i.e., when the number of processors is increased from 2 to 64

Compiler Assisted Leakage Energy Optimization for Clustered VLIW Architectures

Miniaturization of devices and the ensuing decrease in the threshold voltage has led to a substantial increase in the leakage component of the total processor energy consumption. Relatively simpler issue logic and the presence of a large number of function units in the VLIW and the clustered VLIW architectures attribute a large fraction of this leakage energy consumption in the functional units. However, functional units are not fully utilized in the VLIW architectures because of the inherent variations in the ILP of the programs. This underutilization is even more pronounced in the context of clustered VLIW architectures because of the contentions for the limited number of slow intercluster communication channels which lead to many short idle cycles.In the past, some architectural schemes have been proposed to obtain leakage energy bene .ts by aggressively exploiting the idleness of functional units. However, presence of many short idle cycles cause frequent transitions from the active mode to the sleep mode and vice-versa and adversely a ffects the energy benefits of a purely hardware based scheme. In this paper, we propose and evaluate a compiler instruction scheduling algorithm that assist such a hardware based scheme in the context of VLIW and clustered VLIW architectures. The proposed scheme exploits the scheduling slacks of instructions to orchestrate the functional unit mapping with the objective of reducing the number of transitions in functional units thereby keeping them off for a longer duration. The proposed compiler-assisted scheme obtains a further 12% reduction of energy consumption of functional units with negligible performance degradation over a hardware-only scheme for a VLIW architecture. The benefits are 15% and 17% in the context of a 2-clustered and a 4-clustered VLIW architecture respectively. Our test bed uses the Trimaran compiler infrastructure.

Performance Modeling based on Multidimensional Surface Learning for Performance Predictions of Parallel Applications in Non-Dedicated Environments

Modeling the performance behavior of parallel applications to predict the execution times of the applications for larger problem sizes and number of processors has been an active area of research for several years. The existing curve fitting strategies for performance modeling utilize data from experiments that are conducted under uniform loading conditions. Hence the accuracy of these models degrade when the load conditions on the machines and network change. In this paper, we analyze a curve fitting model that attempts to predict execution times for any load conditions that may exist on the systems during application execution. Based on the experiments conducted with the model for a parallel eigenvalue problem, we propose a multi-dimensional curve-fitting model based on rational polynomials for performance predictions of parallel applications in non-dedicated environments. We used the rational polynomial based model to predict execution times for 2 other parallel applications on systems with large load dynamics. In all the cases, the model gave good predictions of execution times with average percentage prediction errors of less than 20%

An operator-splitting finite element method for the efficient parallel solution of multidimensional population balance systems

A finite element method for solving multidimensional population balance systems is proposed where the balance of fluid velocity, temperature and solute partial density is considered as a two-dimensional system and the balance of particle size distribution as a three-dimensional one. The method is based on a dimensional splitting into physical space and internal property variables. In addition, the operator splitting allows to decouple the equations for temperature, solute partial density and particle size distribution. Further, a nodal point based parallel finite element algorithm for multi-dimensional population balance systems is presented. The method is applied to study a crystallization process assuming, for simplicity, a size independent growth rate and neglecting agglomeration and breakage of particles. Simulations for different wall temperatures are performed to show the effect of cooling on the crystal growth. Although the method is described in detail only for the case of d=2 space and s=1 internal property variables it has the potential to be extendable to d+s variables, d=2, 3 and s >= 1. (C) 2011 Elsevier Ltd. All rights reserved.

On pattern oriented software architecture for the grid

Precision, sophistication and economic factors in many areas of scientific research that demand very high magnitude of compute power is the order of the day. Thus advance research in the area of high performance computing is getting inevitable. The basic principle of sharing and collaborative work by geographically separated computers is known by several names such as metacomputing, scalable computing, cluster computing, internet computing and this has today metamorphosed into a new term known as grid computing. This paper gives an overview of grid computing and compares various grid architectures. We show the role that patterns can play in architecting complex systems, and provide a very pragmatic reference to a set of well-engineered patterns that the practicing developer can apply to crafting his or her own specific applications. We are not aware of pattern-oriented approach being applied to develop and deploy a grid. There are many grid frameworks that are built or are in the process of being functional. All these grids differ in some functionality or the other, though the basic principle over which the grids are built is the same. Despite this there are no standard requirements listed for building a grid. The grid being a very complex system, it is mandatory to have a standard Software Architecture Specification (SAS). We attempt to develop the same for use by any grid user or developer. Specifically, we analyze the grid using an object oriented approach and presenting the architecture using UML. This paper will propose the usage of patterns at all levels (analysis. design and architectural) of the grid development.

Language identification using parallel sub-word recognition

Parallel sub-word recognition (PSWR) is a new model that has been proposed for language identification (LID) which does not need elaborate phonetic labeling of the speech data in a foreign language. The new approach performs a front-end tokenization in terms of sub-word units which are designed by automatic segmentation, segment clustering and segment HMM modeling. We develop PSWR based LID in a framework similar to the parallel phone recognition (PPR) approach in the literature. This includes a front-end tokenizer and a back-end language model, for each language to be identified. Considering various combinations of the statistical evaluation scores, it is found that PSWR can perform as well as PPR, even with broad acoustic sub-word tokenization, thus making it an efficient alternative to the PPR system.

Synthesis of configurable architectures for DSP algorithms

ASICs offer the best realization of DSP algorithms in terms of performance, but the cost is prohibitive, especially when the volumes involved are low. However, if the architecture synthesis trajectory for such algorithms is such that the target architecture can be identified as an interconnection of elementary parameterized computational structures, then it is possible to attain a close match, both in terms of performance and power with respect to an ASIC, for any algorithmic parameters of the given algorithm. Such an architecture is weakly programmable (configurable) and can be viewed as an application specific instruction-set processor (ASIP). In this work, we present a methodology to synthesize ASIPs for DSP algorithms.

Relevance of local parallel theory to the linear stability of laminar separation bubbles

Laminar separation bubbles are thought to be highly non-parallel, and hence global stability studies start from this premise. However, experimentalists have always realized that the flow is more parallel than is commonly believed, for pressure-gradient-induced bubbles, and this is why linear parallel stability theory has been successful in describing their early stages of transition. The present experimental/numerical study re-examines this important issue and finds that the base flow in such a separation bubble becomes nearly parallel due to a strong-interaction process between the separated boundary layer and the outer potential flow. The so-called dead-air region or the region of constant pressure is a simple consequence of this strong interaction. We use triple-deck theory to qualitatively explain these features. Next, the implications of global analysis for the linear stability of separation bubbles are considered. In particular we show that in the initial portion of the bubble, where the flow is nearly parallel, local stability analysis is sufficient to capture the essential physics. It appears that the real utility of the global analysis is perhaps in the rear portion of the bubble, where the flow is highly non-parallel, and where the secondary/nonlinear instability stages are likely to dominate the dynamics.