DOCSLIB.ORG

Quality-Based Workload Scaling for Real-Time Streaming Systems

Total Page：16

File Type：pdf, Size：1020Kb

Download full-text PDF Read full-text
Quality-Based Workload Scaling for Real-Time Streaming Systems Procedia Computer Science Procedia Computer Science 101 , 2016 , Pages 323 – 332 YSC 2016. 5th International Young Scientist Conference on Computational Science,Science Quality-based workload scaling for real-time streaming systems Pavel A. Smirnov, Denis Nasonov ITMO University, St.Petersburg, Russia {smirnp, denis_nasonov}@niuitmo.ru Abstract In this paper we propose an idea to scale workload via elastic quality of solution provided by the particular streaming applications. The contribution of this paper consists of quality-based workload scaling model, implementation details for quality assessment mechanism implemented at the top of Apache Storm and experimental evaluation of the proposed model on a synthetic and real-world (medical) examples. Keywords: scaling model, data streaming, elastic workload, quality of service, apache storm, big data 1 Introduction Nowadays huge amounts of data are produced in real time manner. Autonomous vehicles, physical sensors, social-network activities, stock-exchange markets are typical producers of streaming data, which require immediate real-time processing to be actual in a short period of time. Performance optimization for real-time streaming applications is a complex non-trivial process. The goal of profiliration and optimization process is to find an optimal tradeoff between two states of an application: underutilization and overloading. A result of non-optimal performance occurs queues of tuples behind the overloaded operators. Open-source streaming engines like Apache Storm* are smart enough to handle overloads by pausing an emission if downside operators cannot serve it (this feature is called “backpressure’). But if amounts of data from third-party sources remain unchangeable, an overloaded application will accumulate an endless queues of data, which may be lost as a result of crash. While there are at least 5 ways to scale application performance in general (Ahluwalia, 2007), only two of them become popular in field of data streaming: to increase operator’s parallelism and to add extra hardware resources. The literature overview in field of scaling streaming applications outlined that these two patterns are usually applied simultaneously: parallelism increase is reached via allocation of new operators on a newly-added hardware. From economic point of view new hardware resources cause additional costs, so it makes sense for more effective operators’ placement (called “scheduling”), which also impacts on application’s service rate. * http://storm.apache.org/ Peer-review under responsibility of organizing committee of the scientific committee of the 323 5th International Young Scientist Conference on Computational Science © 2016 The Authors. Published by Elsevier B.V. doi: 10.1016/j.procs.2016.11.038 Quality-based workload scaling for real-time streaming systems Pavel A. Smirnov and Denis Nasonov In contrast with scheduling and resource scaling in this paper we propose the idea to scale workload via elastic quality of solution, which a streaming application provides. The contribution of this paper consists of: the quality-based workload scaling model for streaming applications; the implementation details of quality assessment mechanism implemented at the top of Apache Storm; experimental evaluation of the proposed model on the synthetic and real-world (medical) applications. The remainder of the paper is organized as follows: section II presents the literature study; the formal definition of the proposed model is presented in section III; the implementation details of quality assessment mechanism are described in section IV; section V presents the results of experimental evaluation; section VI contains conclusion and plans for the future work with potential use-cases regarding quality-based mass-serving applications. 2 Related works Today a workload scaling patterns for online streaming applications is the actual state-of-art problem. The widely spread open-source streaming engines like Apache Storm†, Samza‡, Flink§, Twitter Heron** and others cannot scale workload automatically, but may do it according to some signals from user. This lead to appearance of several simultaneous research efforts devoted to active and proactive workload scaling techniques. Paper (Xu, Peng, & Gupta, 2016) describes system called Stela, which provides active workload scaling via regular calculation of service rates (authors call it a congestion or ETP-metric) for all operators and increase it for the most ETP-slowest operators by allocating extra instances on newly allocated hardware. The Stela is implemented on top of the Apache Storm platform and automatically captures performance statistics from Storm API. Stela periodically refreshes operators’ ETPs and automatically scales-in and scales-out topologies in Apache Storm. The research (Heinze et al., 2015) proposes online optimization approach, which automatically detects changes in workload pattern, chooses and applies a scaling strategy to minimize number of hosts used for current workload characteristics. Changes of workload patterns are detected by an adaptive windows approach. The approach is implemented on top of FUGU (Heinze et al., n.d.), which is elastic data stream processing engine developed by the authors during previous research efforts. For overloaded hosts the system automatically decides which operators should be unchanged and which should be moved to a newly added resources. The decisions are made according to local and global threshold- based rules (Heinze, Pappalardo, Jerzak, & Fetzer, 2014), which deal with operators’ performance metrics. Authors also propose a QoS-based (Heinze, Jerzak, Hackenbroich, & Fetzer, 2014) approach for workload scaling, where latency is the main QoS constraint, usually strictly declared in service level agreement. In (De Matteis & Mencagli, 2016) authors propose latency latency-aware and energy-efficient scaling strategies with predictive capabilities. The strategies are made via adaptation of the Model Predictive Control – a technique for searching optimal applications’ configurations along a limited prediction horizon. The paper presents models, which describe dependency between QoS variables (latency, energy) and a configuration of the system (parallelism, CPU frequency and etc.). The result of optimization is a reconfiguration trajectory within a strictly-limited prediction horizon. Paper (Hidalgo, Wladdimiro, & Rosas, 2016) is devoted to reactive and predictive resource scaling according to workload demands. Authors apply optimization approach called fission to increase the amount of replicas for fully-loaded operators. The two algorithms are proposed to determine an operator’s state: a short-term and a mid-term state for peaks and patterns detection respectively. The † http://storm.apache.org/ ‡ http://samza.apache.org/ § http://flink.apache.org/ ** http://twitter.github.io/heron/ 324 Quality-based workload scaling for real-time streaming systems Pavel A. Smirnov and Denis Nasonov short-term algorithm classifies operator’s states according to predefined lower and upper thresholds. The mid-term algorithm uses a Markov Chain Model to calculate operators’ transition probabilities. The solution was implemented on top of S4 platform according to MAPE†† model. All the aforementioned papers provide either reactive or predictive workload scaling patterns via modification of a resource set (to add or to remove resource nodes). The major difference between aforementioned papers and our research is in subject for scale: we propose scaling via tuning of quality- sensitive application’s parameters instead of scaling via a resource set. The proposed quality-based scaling provides control for both: performance improvements and a quality loss. To the best of our knowledge there is no research regarding elastic quality for processing results. 3 Quality-based workload scaling In this section we present a formal definition for the proposed quality-based workload scaling model. The model is based on the existing models and extends them with novel formulations. 3.1 Model definition For the model we reuse some elements from third-party streaming application model based queuing theory (Beard & Chamberlain, 2013): ൌሺܸǡܧሻ – an application graph defined by set of vertex ܸ, which define operators and set of edges ܧ, which define flows of data; ɉሺܸ୧ሻ – mean operator’s data arrival rate; Ɋሺܸ୧ሻ – mean operator’s service rate; Here we introduce a matrix of operator’s instances defined as follows: ܹሺܸ୧ሻ=ሼݓଵ ǥݓ௝ሽ. To deal with operator’s performance and quality of it’s results we offer to reuse the performance model from quality-based approach for workflows (Butakov, Nasonov, Svitenkov, & Radice, 2016): ܶ௏௜ ൌ ݂ሺ݀ǡ݌ǡݎሻ – a performance model (calculation time) for operator ܸ௜, which depends on: ݀ = ݀௏௜ – characteristics of an input data (format, size, precision) arrived ݎ a particular resource – ܴא ൌ ௪೔ೕݎ ,௏௜ - one of available parameters’ configurationܲאto operator, ݌ node, where operator’s instance ݓ௜௝ will be launched (note, that ݌ and ݀ are the same for all instances :ݓ௜௝). Also we offer to reuse model of quality function and define for operator ܸ௜ is defined as follows ௏௜ ݍൌܳ ሺܳ௏௜ǡܲ௏௜ሻ, where ܳ௏௜ are quality parameters to measure, ܲ௏௜ ൌሼ݌ଵǡǤǤǤǡ݌௡ሽ parameter’s values, which impact on ܳ௏௜. To declare the operator’s allocation on a particular resource we introduce a schedule function: ܵܿ ൌ ݐ is a scheduling strategy. Theݏ ,is an application graph, ܴ is a resource pool ܩ ݐሻ, whereݏǡܴǡܩሺܨ aggregated elements notation is presented
Load more

Recommended publications
  • Analysis of GPU-Libraries for Rapid Prototyping Database Operations
    Analysis of GPU-Libraries for Rapid Prototyping Database Operations A look into library support for database operations Harish Kumar Harihara Subramanian Bala Gurumurthy Gabriel Campero Durand David Broneske Gunter Saake University of Magdeburg Magdeburg, Germany fi[email protected] Abstract—Using GPUs for query processing is still an ongoing Usually, library operators for GPUs are either written by research in the database community due to the increasing hardware experts [12] or are available out of the box by device heterogeneity of GPUs and their capabilities (e.g., their newest vendors [13]. Overall, we found more than 40 libraries for selling point: tensor cores). Hence, many researchers develop optimal operator implementations for a specific device generation GPUs each packing a set of operators commonly used in one involving tedious operator tuning by hand. On the other hand, or more domains. The benefits of those libraries is that they are there is a growing availability of GPU libraries that provide constantly updated and tested to support newer GPU versions optimized operators for manifold applications. However, the and their predefined interfaces offer high portability as well as question arises how mature these libraries are and whether they faster development time compared to handwritten operators. are fit to replace handwritten operator implementations not only w.r.t. implementation effort and portability, but also in terms of This makes them a perfect match for many commercial performance. database systems, which can rely on GPU libraries to imple- In this paper, we investigate various general-purpose libraries ment well performing database operators. Some example for that are both portable and easy to use for arbitrary GPUs such systems are: SQreamDB using Thrust [14], BlazingDB in order to test their production readiness on the example of using cuDF [15], Brytlyt using the Torch library [16].
    [Show full text]
  • Concurrent Cilk: Lazy Promotion from Tasks to Threads in C/C++
    Concurrent Cilk: Lazy Promotion from Tasks to Threads in C/C++ Christopher S. Zakian, Timothy A. K. Zakian Abhishek Kulkarni, Buddhika Chamith, and Ryan R. Newton Indiana University - Bloomington, fczakian, tzakian, adkulkar, budkahaw, [email protected] Abstract. Library and language support for scheduling non-blocking tasks has greatly improved, as have lightweight (user) threading packages. How- ever, there is a significant gap between the two developments. In previous work|and in today's software packages|lightweight thread creation incurs much larger overheads than tasking libraries, even on tasks that end up never blocking. This limitation can be removed. To that end, we describe an extension to the Intel Cilk Plus runtime system, Concurrent Cilk, where tasks are lazily promoted to threads. Concurrent Cilk removes the overhead of thread creation on threads which end up calling no blocking operations, and is the first system to do so for C/C++ with legacy support (standard calling conventions and stack representations). We demonstrate that Concurrent Cilk adds negligible overhead to existing Cilk programs, while its promoted threads remain more efficient than OS threads in terms of context-switch overhead and blocking communication. Further, it enables development of blocking data structures that create non-fork-join dependence graphs|which can expose more parallelism, and better supports data-driven computations waiting on results from remote devices. 1 Introduction Both task-parallelism [1, 11, 13, 15] and lightweight threading [20] libraries have become popular for different kinds of applications. The key difference between a task and a thread is that threads may block|for example when performing IO|and then resume again.
    [Show full text]
  • Scalable and High Performance MPI Design for Very Large
    SCALABLE AND HIGH-PERFORMANCE MPI DESIGN FOR VERY LARGE INFINIBAND CLUSTERS DISSERTATION Presented in Partial Fulfillment of the Requirements for the Degree Doctor of Philosophy in the Graduate School of The Ohio State University By Sayantan Sur, B. Tech ***** The Ohio State University 2007 Dissertation Committee: Approved by Prof. D. K. Panda, Adviser Prof. P. Sadayappan Adviser Prof. S. Parthasarathy Graduate Program in Computer Science and Engineering c Copyright by Sayantan Sur 2007 ABSTRACT In the past decade, rapid advances have taken place in the field of computer and network design enabling us to connect thousands of computers together to form high-performance clusters. These clusters are used to solve computationally challenging scientific problems. The Message Passing Interface (MPI) is a popular model to write applications for these clusters. There are a vast array of scientific applications which use MPI on clusters. As the applications operate on larger and more complex data, the size of the compute clusters is scaling higher and higher. Thus, in order to enable the best performance to these scientific applications, it is very critical for the design of the MPI libraries be extremely scalable and high-performance. InfiniBand is a cluster interconnect which is based on open-standards and gaining rapid acceptance. This dissertation presents novel designs based on the new features offered by InfiniBand, in order to design scalable and high-performance MPI libraries for large-scale clusters with tens-of-thousands of nodes. Methods developed in this dissertation have been applied towards reduction in overall resource consumption, increased overlap of computa- tion and communication, improved performance of collective operations and finally designing application-level benchmarks to make efficient use of modern networking technology.
    [Show full text]
  • DNA Scaffolds Enable Efficient and Tunable Functionalization of Biomaterials for Immune Cell Modulation
    ARTICLES https://doi.org/10.1038/s41565-020-00813-z DNA scaffolds enable efficient and tunable functionalization of biomaterials for immune cell modulation Xiao Huang 1,2, Jasper Z. Williams 2,3, Ryan Chang1, Zhongbo Li1, Cassandra E. Burnett4, Rogelio Hernandez-Lopez2,3, Initha Setiady1, Eric Gai1, David M. Patterson5, Wei Yu2,3, Kole T. Roybal2,4, Wendell A. Lim2,3 ✉ and Tejal A. Desai 1,2 ✉ Biomaterials can improve the safety and presentation of therapeutic agents for effective immunotherapy, and a high level of control over surface functionalization is essential for immune cell modulation. Here, we developed biocompatible immune cell-engaging particles (ICEp) that use synthetic short DNA as scaffolds for efficient and tunable protein loading. To improve the safety of chimeric antigen receptor (CAR) T cell therapies, micrometre-sized ICEp were injected intratumorally to present a priming signal for systemically administered AND-gate CAR-T cells. Locally retained ICEp presenting a high density of priming antigens activated CAR T cells, driving local tumour clearance while sparing uninjected tumours in immunodeficient mice. The ratiometric control of costimulatory ligands (anti-CD3 and anti-CD28 antibodies) and the surface presentation of a cytokine (IL-2) on ICEp were shown to substantially impact human primary T cell activation phenotypes. This modular and versatile bio- material functionalization platform can provide new opportunities for immunotherapies. mmune cell therapies have shown great potential for cancer the specific antibody27, and therefore a high and controllable den- treatment, but still face major challenges of efficacy and safety sity of surface biomolecules could allow for precise control of ligand for therapeutic applications1–3, for example, chimeric antigen binding and cell modulation.
    [Show full text]
  • A Cilk Implementation of LTE Base-Station Up- Link on the Tilepro64 Processor
    A Cilk Implementation of LTE Base-station up- link on the TILEPro64 Processor Master of Science Thesis in the Programme of Integrated Electronic System Design HAO LI Chalmers University of Technology Department of Computer Science and Engineering Goteborg,¨ Sweden, November 2011 The Author grants to Chalmers University of Technology and University of Gothen- burg the non-exclusive right to publish the Work electronically and in a non-commercial purpose make it accessible on the Internet. The Author warrants that he/she is the au- thor to the Work, and warrants that the Work does not contain text, pictures or other material that violates copyright law. The Author shall, when transferring the rights of the Work to a third party (for ex- ample a publisher or a company), acknowledge the third party about this agreement. If the Author has signed a copyright agreement with a third party regarding the Work, the Author warrants hereby that he/she has obtained any necessary permission from this third party to let Chalmers University of Technology and University of Gothenburg store the Work electronically and make it accessible on the Internet. A Cilk Implementation of LTE Base-station uplink on the TILEPro64 Processor HAO LI © HAO LI, November 2011 Examiner: Sally A. McKee Chalmers University of Technology Department of Computer Science and Engineering SE-412 96 Goteborg¨ Sweden Telephone +46 (0)31-255 1668 Supervisor: Magnus Sjalander¨ Chalmers University of Technology Department of Computer Science and Engineering SE-412 96 Goteborg¨ Sweden Telephone +46 (0)31-772 1075 Cover: A Cilk Implementation of LTE Base-station uplink on the TILEPro64 Processor.
    [Show full text]
  • Library for Handling Asynchronous Events in C++
    Masaryk University Faculty of Informatics Library for Handling Asynchronous Events in C++ Bachelor’s Thesis Branislav Ševc Brno, Spring 2019 Declaration Hereby I declare that this paper is my original authorial work, which I have worked out on my own. All sources, references, and literature used or excerpted during elaboration of this work are properly cited and listed in complete reference to the due source. Branislav Ševc Advisor: Mgr. Jan Mrázek i Acknowledgements I would like to thank my supervisor, Jan Mrázek, for his valuable guid- ance and highly appreciated help with avoiding death traps during the design and implementation of this project. ii Abstract The subject of this work is design, implementation and documentation of a library for the C++ programming language, which aims to sim- plify asynchronous event-driven programming. The Reactive Blocks Library (RBL) allows users to define program as a graph of intercon- nected function blocks which control message flow inside the graph. The benefits include decoupling of program’s logic and the method of execution, with emphasis on increased readability of program’s logical behavior through understanding of its reactive components. This thesis focuses on explaining the programming model, then proceeds to defend design choices and to outline the implementa- tion layers. A brief analysis of current competing solutions and their comparison with RBL, along with overhead benchmarks of RBL’s abstractions over pure C++ approach is included. iii Keywords C++ Programming Language, Inversion
    [Show full text]
  • Intel Threading Building Blocks
    Praise for Intel Threading Building Blocks “The Age of Serial Computing is over. With the advent of multi-core processors, parallel- computing technology that was once relegated to universities and research labs is now emerging as mainstream. Intel Threading Building Blocks updates and greatly expands the ‘work-stealing’ technology pioneered by the MIT Cilk system of 15 years ago, providing a modern industrial-strength C++ library for concurrent programming. “Not only does this book offer an excellent introduction to the library, it furnishes novices and experts alike with a clear and accessible discussion of the complexities of concurrency.” — Charles E. Leiserson, MIT Computer Science and Artificial Intelligence Laboratory “We used to say make it right, then make it fast. We can’t do that anymore. TBB lets us design for correctness and speed up front for Maya. This book shows you how to extract the most benefit from using TBB in your code.” — Martin Watt, Senior Software Engineer, Autodesk “TBB promises to change how parallel programming is done in C++. This book will be extremely useful to any C++ programmer. With this book, James achieves two important goals: • Presents an excellent introduction to parallel programming, illustrating the most com- mon parallel programming patterns and the forces governing their use. • Documents the Threading Building Blocks C++ library—a library that provides generic algorithms for these patterns. “TBB incorporates many of the best ideas that researchers in object-oriented parallel computing developed in the last two decades.” — Marc Snir, Head of the Computer Science Department, University of Illinois at Urbana-Champaign “This book was my first introduction to Intel Threading Building Blocks.
    [Show full text]
  • Lecture 15-16: Parallel Programming with Cilk
    Lecture 15-16: Parallel Programming with Cilk Concurrent and Mul;core Programming Department of Computer Science and Engineering Yonghong Yan [email protected] www.secs.oakland.edu/~yan 1 Topics (Part 1) • IntroducAon • Principles of parallel algorithm design (Chapter 3) • Programming on shared memory system (Chapter 7) – OpenMP – Cilk/Cilkplus – PThread, mutual exclusion, locks, synchroniza;ons • Analysis of parallel program execuAons (Chapter 5) – Performance Metrics for Parallel Systems • Execuon Time, Overhead, Speedup, Efficiency, Cost – Scalability of Parallel Systems – Use of performance tools 2 Shared Memory Systems: Mul;core and Mul;socket Systems • a 3 Threading on Shared Memory Systems • Employ parallelism to compute on shared data – boost performance on a fixed memory footprint (strong scaling) • Useful for hiding latency – e.g. latency due to I/O, memory latency, communicaon latency • Useful for scheduling and load balancing – especially for dynamic concurrency • Relavely easy to program – easier than message-passing? you be the judge! 4 Programming Models on Shared Memory System • Library-based models – All data are shared – Pthreads, Intel Threading Building Blocks, Java Concurrency, Boost, Microso\ .Net Task Parallel Library • DirecAve-based models, e.g., OpenMP – shared and private data – pragma syntax simplifies thread creaon and synchronizaon • Programming languages – CilkPlus (Intel, GCC), and MIT Cilk – CUDA (NVIDIA) – OpenCL 5 Toward Standard Threading for C/C++ At last month's meeting of the C standard committee, WG14 decided to form a study group to produce a proposal for language extensions for C to simplify parallel programming. This proposal is expected to combine the best ideas from Cilk and OpenMP, two of the most widely-used and well-established parallel language extensions for the C language family.
    [Show full text]
  • SYCL – a Gentle Introduction
    SYCL – A gentle Introduction Thomas Applencourt - [email protected], Abhishek Bagusetty Argonne Leadership Computing Facility Argonne National Laboratory 9700 S. Cass Ave Argonne, IL 60349 Table of contents 1. Introduction 2. DPCPP ecosystem 3. Theory Context And Queue Unified Shared Memory 4. Kernel Submission 5. Buffer innovation 6. Conclusion 1/28 Introduction What programming model to target Accelerator? • CUDA1 / HIP2 / OpenCL3 • OpenMP (pragma based) • Kokkos, raja, OCCA (high level, abstraction layer, academic project) • SYCL (high level) / DPCPP4 • Parallel STL5 1Compute Unified Device Architecture 2Heterogeneous-Compute Interface 3Open Computing Language 4Data Parallel C++ 5SYCL implementation exist https://github.com/oneapi-src/oneDPL 2/28 What is SYCL™? 1. Target C++ programmers (template, lambda) • No language extension • No pragmas • No attribute 2. Borrow lot of concept from battle tested OpenCL (platform, device, work-group, range) 3. Single Source (two compilation pass) 4. Implicit or Explicit data-transfer 5. SYCL is a Specification developed by the Khronos Group (OpenCL, SPIR, Vulkan, OpenGL) • The current stable SYCL specification is SYCL2020 3/28 SYCL Implementation 6 6Credit: Khronos groups (https://www.khronos.org/sycl/) 4/28 What is DPCPP? • Intel implementation of SYCL • The name of the SYCL-aware Intel compiler7 who is packaged with Intel OneAPI SDK. • Intel SYCL compiler is open source and based on LLVM https://github.com/intel/llvm/. This is what is installed on ThetaGPU, hence the compiler will be named 8 clang++ . 7So you don’t need to pass -fsycl 8I know marketing is confusing... 5/28 How to install SYCL: Example with Intel implementation • Intel implementation work with Intel and NVDIA Hardware 1.
    [Show full text]
  • High Performance Integration of Data Parallel File Systems and Computing
    HIGH PERFORMANCE INTEGRATION OF DATA PARALLEL FILE SYSTEMS AND COMPUTING: OPTIMIZING MAPREDUCE Zhenhua Guo Submitted to the faculty of the University Graduate School in partial fulfillment of the requirements for the degree Doctor of Philosophy in the Department of Computer Science Indiana University August 2012 Accepted by the Graduate Faculty, Indiana University, in partial fulfillment of the require- ments of the degree of Doctor of Philosophy. Doctoral Geoffrey Fox, Ph.D. Committee (Principal Advisor) Judy Qiu, Ph.D. Minaxi Gupta, Ph.D. David Leake, Ph.D. ii Copyright c 2012 Zhenhua Guo ALL RIGHTS RESERVED iii I dedicate this dissertation to my parents and my wife Mo. iv Acknowledgements First and foremost, I owe my sincerest gratitude to my advisor Prof. Geoffrey Fox. Throughout my Ph.D. research, he guided me into the research field of distributed systems; his insightful advice inspires me to identify the challenging research problems I am interested in; and his generous intelligence support is critical for me to tackle difficult research issues one after another. During the course of working with him, I learned how to become a professional researcher. I would like to thank my entire research committee: Dr. Judy Qiu, Prof. Minaxi Gupta, and Prof. David Leake. I am greatly indebted to them for their professional guidance, generous support, and valuable suggestions that were given throughout this research work. I am grateful to Dr. Judy Qiu for offering me the opportunities to participate into closely related projects. As a result, I obtained deeper understanding of related systems including Dryad and Twister, and could better position my research in the big picture of the whole research area.
    [Show full text]
  • Trisycl Open Source C++17 & Openmp-Based Opencl SYCL Prototype
    triSYCL Open Source C++17 & OpenMP-based OpenCL SYCL prototype Ronan Keryell Khronos OpenCL SYCL committee 05/12/2015 — IWOCL 2015 SYCL Tutorial • I OpenCL SYCL committee work... • Weekly telephone meeting • Define new ways for modern heterogeneous computing with C++ I Single source host + kernel I Replace specific syntax by pure C++ abstractions • Write SYCL specifications • Write SYCL conformance test • Communication & evangelism triSYCL — Open Source C++17 & OpenMP-based OpenCL SYCL prototype IWOCL 2015 2 / 23 • I SYCL relies on advanced C++ • Latest C++11, C++14... • Metaprogramming • Implicit type conversion • ... Difficult to know what is feasible or even correct... ;• Need a prototype to experiment with the concepts • Double-check the specification • Test the examples Same issue with C++ standard and GCC or Clang/LLVM triSYCL — Open Source C++17 & OpenMP-based OpenCL SYCL prototype IWOCL 2015 3 / 23 • I Solving the meta-problem • SYCL specification I Includes header files descriptions I Beautified .hpp I Tables describing the semantics of classes and methods • Generate Doxygen-like web pages Generate parts of specification and documentation from a reference implementation ; triSYCL — Open Source C++17 & OpenMP-based OpenCL SYCL prototype IWOCL 2015 4 / 23 • I triSYCL (I) • Started in April 2014 as a side project • Open Source for community purpose & dissemination • Pure C++ implementation I DSEL (Domain Specific Embedded Language) I Rely on STL & Boost for zen style I Use OpenMP 3.1 to leverage CPU parallelism I No compiler cannot generate kernels on GPU yet • Use Doxygen to generate I Documentation; of triSYCL implementation itself with implementation details I SYCL API by hiding implementation details with macros & Doxygen configuration • Python script to generate parts of LaTeX specification from Doxygen LaTeX output triSYCL — Open Source C++17 & OpenMP-based OpenCL SYCL prototype IWOCL 2015 5 / 23 • I Automatic generation of SYCL specification is a failure..
    [Show full text]
  • Optimizing Apache Spark* to Maximize Workload Throughput
    TECHNOLOGY BRIEF Data Center Big Data Processing Optimizing Apache Spark* to Maximize Workload Throughput Apache Spark* throughput doubled and runtime was reduced by 40% with Intel® Optane™ SSD DC P4800X and Intel® Memory Drive Technology. Executive Summary Apache Spark* is a popular data processing engine designed to execute advanced analytics on very large data sets which are common in today’s enterprise use cases involving Cloud-based services, IoT, and Machine Learning. Spark implements a general-purpose clustered computing framework that can ingest and process real-time streams of very large data, enabling instantaneous event- and exception- handling, analytics, and decision-making for responsive user interaction. To enable Spark’s high performance for different workloads (e.g. machine-learning applications), in-memory data storage capabilities are built right in. Consequently, Spark significantly outperforms the alternative big data processing technologies. However, Spark’s in-memory capabilities are limited by the memory available in the server; it is common for computing resources to be idle during the execution of a Spark job, even though the system’s memory is saturated. To mitigate this limitation, Spark’s distributed architecture can run on a cluster of nodes, thus taking advantage of the memory available across all nodes. While employing additional nodes would solve the server DRAM capacity problem, it does so at an increased cost. DRAM is not only expensive, it also requires that the operator provision additional servers in order to procure the additional memory. Intel® Memory Drive Technology is a software-defined memory (SDM) technology, which combined with an Intel® Optane™ SSD, expands the system’s memory.
    [Show full text]