Modern Multi Research Articles

Well-structured futures and cache locality

In fork-join parallelism , a sequential program is split into a directed acyclic graph of tasks linked by directed dependency edges, and the tasks are executed, possibly in parallel, in an order consistent with their dependencies. A popular and effective way to extend fork-join parallelism is to allow threads to create {futures . A thread creates a future to hold the results of a computation, which may or may not be executed in parallel. That result is returned when some thread touches that future, blocking if necessary until the result is ready. Recent research has shown that while futures can, of course, enhance parallelism in a structured way, they can have a deleterious effect on cache locality. In the worst case, futures can incur Ω(P T∞ + t T∞) deviations, which implies Ω(C P T∞ + C t T∞) additional cache misses, where C is the number of cache lines, P is the number of processors, t is the number of touches, and T∞ is the computation span . Since cache locality has a large impact on software performance on modern multicores, this result is troubling. In this paper, however, we show that if futures are used in a simple, disciplined way, then the situation is much better: if each future is touched only once, either by the thread that created it, or by a later descendant of the thread that created it, then parallel executions with work stealing can incur at most O(C P T 2 ∞) additional cache misses, a substantial improvement. This structured use of futures is characteristic of many (but not all) parallel applications.

Bathymetry and geographical regionalization of Brepollen (Hornsund, Spitsbergen) based on bathymetric profiles interpolations

Abstract Determination of High Arctic regions bathymetry is strictly dependent from weather and ice mass quantity. Due to safety, it is often necessary to use a small boat to study fjords area, especially close to glaciers with unknown bathymetry. This precludes the use of modern multi−beam echosounders, and so traditional single−beam echosounders have been used for bathymetry profiling. Adequate interpolation techniques were determined for the most probable morphological formations in−between bathymetric profiles. Choosing the most accurate interpolation method allows for the determination of geographical regionalisation of submarine elevations of the Brepollen area (inner part of Hornsund, Spitsbergen). It has also been found that bathymetric interpolations should be performed on averaged grid values, rather than individual records. The Ordinary Kriging Method was identified as the most adequate for interpolations and was compared with multi beam scan− ning, which was possible to make due to a previously modelled single beam interpolation map. In total, eight geographical units were separated in Brepollen, based on the bathy− metry, slope and aspect maps. Presented results provide a truly new image of the area, which allow for further understanding of past and present processes in the High Arctic.

Can traditional programming bridge the Ninja performance gap for parallel computing applications?

Current processor trends of integrating more cores with wider SIMD units, along with a deeper and complex memory hierarchy, have made it increasingly more challenging to extract performance from applications. It is believed by some that traditional approaches to programming do not apply to these modern processors and hence radical new languages must be discovered. In this paper, we question this thinking and offer evidence in support of traditional programming methods and the performance-vs-programming effort effectiveness of common multi-core processors and upcoming many-core architectures in delivering significant speedup, and close-to-optimal performance for commonly used parallel computing workloads. We first quantify the extent of the " Ninja gap ", which is the performance gap between naively written C/C++ code that is parallelism unaware (often serial) and best-optimized code on modern multi-/many-core processors. Using a set of representative throughput computing benchmarks, we show that there is an average Ninja gap of 24X (up to 53X ) for a recent 6-core Intel® Core™ i7 X980 Westmere CPU, and that this gap if left unaddressed will inevitably increase. We show how a set of well-known algorithmic changes coupled with advancements in modern compiler technology can bring down the Ninja gap to an average of just 1.3X . These changes typically require low programming effort, as compared to the very high effort in producing Ninja code. We also discuss hardware support for programmability that can reduce the impact of these changes and even further increase programmer productivity. We show equally encouraging results for the upcoming Intel® Many Integrated Core architecture (Intel® MIC) which has more cores and wider SIMD. We thus demonstrate that we can contain the otherwise uncontrolled growth of the Ninja gap and offer a more stable and predictable performance growth over future architectures, offering strong evidence that radical language changes are not required.

Computational geometry in the parallel external memory model

Continued advances in VLSI scaling combined with unsustainable power consumption of frequency scaling resulted in parallel processors having become mainstream as demonstrated by modern multicores. Current off-the-shelf processors already contain up to 16 cores and the prototypes boast up to 48 cores on a single chip.

UML Profile and Extensions for Complex Approval Systems with Complementary Levels of Abstraction

Extending the Unified Modelling Language (UML) version 2.0 with a profile and stereotypes that allow the modelling of an approval system with multiple levels, each with inherently dualistic complementarity, allows the high-fidelity characterization of the stacked levels of a real world approval hierarchy. The objective was to extend UML to model the complex, emergent and multi-level decision making which occurs within modern multi- disciplinary projects. At any level of abstraction, the logical and concrete processes of that level allow a consideration of local factors and decisions, generating and establishing a qualitative conclusion as the result. The high degree of certainly in the result creates the absoluteness of the qualitative conclusion, which can then be fed up or down one level of abstraction in order to take part in the local decision making at that level. The profile was applied to ground-breaking and ongoing engineering investigation concerning the expansion of a student busing system as it is proposed to be integrated into a city-wide busing system, where the student busing system can be considered to be a sub-system to the city-wide busing system, but in reality it is a self-standing, independent and complete system in its own right.

Effective Implementation of DGEMM on Modern Multicore CPU

In this paper we will present a detailed study on tuning double-precision matrix-matrix multiplication (DGEMM) on the Intel Xeon E5-2680 CPU. We selected an optimal algorithm from the instruction set perspective as well software tools optimized for Intel Advance Vector Extensions (AVX). Our optimizations included the use of vector memory operations, and AVX instructions. Our proposed algorithm achieves a performance improvement of 33% compared to the latest results achieved using the Intel Math Kernel Library DGEMM subroutine.

Determination of optimal ordering quantity and reduction of bullwhip effect in a multistage supply chain using genetic algorithm

Allocating optimal ordering quantity and mitigation of bullwhip effect is one of the challenging parts in a modern multi echelon supply chain system. Genetic algorithm is used in this research to reduce the bullwhip effect and to determine optimal ordering quantity in a multistage supply chain consisting of six members. Real demand data of a manufacturing company has been used here to conduct the analyses. This research also pinpointed that genetic algorithm can be applied to reduce the cost of total supply chain. To calculate the total cost, five different costs with varying unit holding cost, in inventory cost is used for each of the members of supply chain. This paper also examines the importance of the limit set of the chromosomes of genetic algorithm and concluded that the lower range chromosomes provide a better result by reducing the total supply chain cost more than the range set in higher limit.

A Cost-Effective Hardware Approach for Measuring Power Consumption of Modern Multi-Core Processors

Multiple processor cores are built within a chip by advanced VLSI technology. With the decreasing prices, multi-core processors are widely deployed in both server and desktop systems. The workload of multi-threaded applications could be separated to different cores by multiple threads, such that application threads can run concurrently to maximize overall execution speed of the applications. Moreover, for the green trend of computing nowadays, most of modern multi-core processors have a functionality of dynamic frequency turning. The power-level tuning techniques are based on Dynamic Voltage and Frequency Scaling (DVFS). In order to evaluate the performance of various power-saving approaches, an appropriate technique to measure the power consumption of multi-core processors is important. However, most of approaches estimate CPU power consumption only from CMOS power consumption data and CPU frequency. These approaches only estimate the dynamic power consumption of multi-core processors, the static power consumption is not be included. In this study, a hardware approach for the power consumption measurement of multi-core processors is proposed. Thus the power consumption of a CPU could be measured precisely, and the performance of CPU power-saving approaches can be evaluated well.

A simulation suite for Lattice-Boltzmann based real-time CFD applications exploiting multi-level parallelism on modern multi- and many-core architectures

We present a software approach to hardware-oriented numerics which builds upon an augmented, previously published set of open-source libraries facilitating portable code development and optimisation on a wide range of modern computer architectures. In order to maximise efficiency, we exploit all levels of parallelism, including vectorisation within CPU cores, the Cell BE and GPUs, shared memory thread-level parallelism between cores, and parallelism between heterogeneous distributed memory resources in clusters. To evaluate and validate our approach, we implement a collection of modular building blocks for the easy and fast assembly and development of CFD applications based on the shallow water equations: We combine the Lattice-Boltzmann method with fluid-structure interaction techniques in order to achieve real-time simulations targeting interactive virtual environments. Our results demonstrate that recent multi-core CPUs outperform the Cell BE, while GPUs are significantly faster than conventional multi-threaded SSE code. In addition, we verify good scalability properties of our application on small clusters.

Optimization Strategy of Top-Down Join Enumeration on Modern Multi-Core CPUs

Most contemporary database systems query optimizers exploit System-R’s bottom-up dynamic programming method (DP) to find the optimal query execution plan (QEP) without evaluating redundant sub-plans. The distinguished exceptions are Volcano/Cascades using transforms to generate new plans according to a top-down approach. As recent research has revealed, bottom-up dynamic programming can improve performance with respect to the shape of the join graph and parallelism. However top-down join enumeration dynamic programming method can derive upper bounds for the costs of the plans it generates which is not available to typical bottom-up DP method. In this paper, we propose a comprehensive and practical framework for parallelizing top-down dynamic programming query optimization with complex non-inner join in the multi-core processor architecture, referred as PTDhyp. We have implemented such a search strategy and experimental results show that can improve optimization time effective compared to known existing algorithms.

Research of Collectives Optimization on Modern Multicore Clusters

With the rapid increase in HPC computing requirement,the multicore is widely deployed in HPC systems.To keep the efficiency of application in large scale systems,it is very important keep up the balance of communication to computation,thus multicore brings more requirement for communication systems.As collective communication is an important part in communication systems and is critical for the whole systems,thus it is important to research on the impacts of multicore environment on collective performance.This paper first analyzes how multicore impacts on collective communication.It is found out that multicore SMP clusters brings two conflict impacts,it not only has faster intra-socket communication path which can speed up the performance of collective communications,but also it brings memory/cache contention which might degrade the communication performance.Based on these aspects,this paper proposes multicore-aware collecitves optimization techniques,which includes:hierarchy-aware algorithms,limited-concurrency,NUMA-aware algorithms and cache-friendly optimization.These optimization methods are implemented in MPI_Barrier,MPI_Bcast and MPI_Alltoall.Experiments show that the proposed algorithms increase the performance and scalability of collective communication.

A fast‐adaptive composite grid algorithm for solving the free‐space Poisson problem on the cell broadband engine

AbstractFast solvers for Poisson's equation with boundary conditions at infinity are an important building block for molecular dynamics. One issue that arises when this equation is solved numerically is the infinite size of the domain. This prevents a direct solution so that other concepts have to be considered. Within this paper a method is discussed that employs hierarchically coarsened grids to overcome this problem. Special attention has to be paid to the discretization at the grid interfaces. A finite volume approach is used for the same. The resulting set of linear equations is solved using a fast‐adaptive composite grid algorithm. Emphasis is put on the implementation of the method on the STI cell broadband engine, a modern multi core processor, that is powerful in floating point operations and memory bandwidth. Code optimization techniques are applied as well as parallelization of the code to get maximum performance on this processor. For validation of the performance test runs are executed and the runtime is analyzed in detail. Copyright © 2010 John Wiley & Sons, Ltd.

RAID Architecture with Correction of Corrupted Data in Faulty Disk Blocks

Disk capacities and processor performances have increased dramatically ever since. With rising storage space the probability of failures gets higher. Reliability of storage systems is achieved by adding extra disks for redundancy, like RAID-Systems or separate backup space in general. These systems cover the case when disks fail but do not recognize corrupted data in faulty blocks. Especially new storage systems like Solid State Drives are more vulnerable to corrupted data as cells are “aging” over time. We propose to add error detection and correction of data to a RAID-system without increasing the amount of space needed to store redundancy information compared to the common implementation RAID 6. To overcome higher computation complexity the implementation uses parallel execution paths available in modern Multicore and Multiprocessor systems.

MUDD

Today's business intelligence systems consist of hundreds of processors with disk subsystems able to handle multiple Giga-bytes of IO-bandwidth. These systems usually contain terabytes of data. Evaluating database system performance of such systems often requires generating synthetic data with well defined statistical properties. To simulate different scenarios, it is important to vary statistical properties including row counts of tables. Foremost, in order to analyze large scale systems, data generators need to be able to produce hundreds of terabytes of data in a timely fashion. In this paper we present MUDD, a multi-dimensional data generator. Originally designed for TPC-DS, a decision support benchmark being developed by the TPC, MUDD is able to generate up to 100 Terabyte of flat file data in hours, utilizing modern multi processor architectures, including clusters. Its novel design separates data generation algorithms from data distribution definitions, enabling users to adjust their workload to individual needs and different scenarios.

Motivations and Performance Conditions for Ethnic Entrepreneurship

Ethnic entrepreneurship has become a popular concept in a modern multi–cultural society. This paper seeks to offer an overview of the potential of ethnic entrepreneurship for solving inter alia the structural unemployment problems of ethnic groups in cities. There is a general lack of empirical evidence on this issue in the current literature. The present paper aims to fill this gap by addressing in particular the critical success conditions for ethnic entrepreneurs. The focus of the research is on variations in success across three ethnic groups in the Amsterdam area. By means of structured personal interviews with many ethnic entrepreneurs, a systematic qualitative data base was created. The paper sets out to identify empirically the driving forces for business success, such as education or the role of informal networks. The explanatory framework deployed for the identification of these qualitative success factors for distinct ethnic groups is based on a particular, recently developed artificial intelligence method, viz. rough set analysis. This multidimensional classification approach appears to be able to identify various important factors for the motivation and performance of ethnic enterprises. Two major findings emerge from this investigation: (i) performance conditions vary across ethnic groups, and (ii) informal networks are crucial for business success.