Intel® Parallel Studio and Advisor

Slides:



Advertisements
Similar presentations
Module 13: Performance Tuning. Overview Performance tuning methodologies Instance level Database level Application level Overview of tools and techniques.
Advertisements

Intel Parallel Advisor Workflow David Valentine Computer Science Slippery Rock University.
Intel® performance analyze tools Nikita Panov Idrisov Renat.
The Path to Multi-core Tools Paul Petersen. Multi-coreToolsThePathTo 2 Outline Motivation Where are we now What is easy to do next What is missing.
Silberschatz, Galvin and Gagne ©2013 Operating System Concepts Essentials – 2 nd Edition Chapter 4: Threads.
Microsoft ® Official Course Monitoring and Troubleshooting Custom SharePoint Solutions SharePoint Practice Microsoft SharePoint 2013.
ORIGINAL AUTHOR JAMES REINDERS, INTEL PRESENTED BY ADITYA AMBARDEKAR Overview for Intel Xeon Processors and Intel Xeon Phi coprocessors.
SEC(R) 2008 Intel® Concurrent Collections for C++ - a model for parallel programming Nikolay Kurtov Software and Services.
Multi-core Programming Thread Profiler. 2 Tuning Threaded Code: Intel® Thread Profiler for Explicit Threads Topics Look at Intel® Thread Profiler features.
OpenMP in a Heterogeneous World Ayodunni Aribuki Advisor: Dr. Barbara Chapman HPCTools Group University of Houston.
WORK ON CLUSTER HYBRILIT E. Aleksandrov 1, D. Belyakov 1, M. Matveev 1, M. Vala 1,2 1 Joint Institute for nuclear research, LIT, Russia 2 Institute for.
Welcome to the Power of 64-bit Computing …now available on your desktop! © 1998, 1999 Compaq Computer Corporation.
國立台灣大學 資訊工程學系 Chapter 4: Threads. 資工系網媒所 NEWS 實驗室 Objectives To introduce the notion of a thread — a fundamental unit of CPU utilization that forms the.
ICOM 5995: Performance Instrumentation and Visualization for High Performance Computer Systems Lecture 7 October 16, 2002 Nayda G. Santiago.
Analyzing parallel programs with Pin Moshe Bach, Mark Charney, Robert Cohn, Elena Demikhovsky, Tevi Devor, Kim Hazelwood, Aamer Jaleel, Chi- Keung Luk,
Database-Driven Web Sites, Second Edition1 Chapter 5 WEB SERVERS.
Multi-core Programming Threading Methodology. 2 Topics A Generic Development Cycle.
Copyright © 2002, Intel Corporation. All rights reserved. *Other brands and names are the property of their respective owners
Debugging parallel programs. Breakpoint debugging Probably the most widely familiar method of debugging programs is breakpoint debugging. In this method,
1 How to do Multithreading First step: Sampling and Hotspot hunting Myongji University Sugwon Hong 1.
Silberschatz, Galvin and Gagne ©2013 Operating System Concepts – 9 th Edition Chapter 4: Threads.
Gauss Students’ Views on Multicore Processors Group members: Yu Yang (presenter), Xiaofang Chen, Subodh Sharma, Sarvani Vakkalanka, Anh Vo, Michael DeLisi,
Integrating and Extending Workflow 8 AA301 Carl Sykes Ed Heaney.
Tuning Threaded Code with Intel® Parallel Amplifier.
Visual Programming Borland Delphi. Developing Applications Borland Delphi is an object-oriented, visual programming environment to develop 32-bit applications.
QNX® Momentics® Development Suite Tools for Building, Debugging and Optimizing Embedded Systems.
Introduction to Performance Tuning Chia-heng Tu PAS Lab Summer Workshop 2009 June 30,
PERFORMANCE OF THE OPENMP AND MPI IMPLEMENTATIONS ON ULTRASPARC SYSTEM Abstract Programmers and developers interested in utilizing parallel programming.
Parallel OpenFOAM CFD Performance Studies Student: Adi Farshteindiker Advisors: Dr. Guy Tel-Zur,Prof. Shlomi Dolev The Department of Computer Science Faculty.
Lecture 5. Example for periority The average waiting time : = 41/5= 8.2.
Chapter Goals Describe the application development process and the role of methodologies, models, and tools Compare and contrast programming language generations.
Chapter 10 Application Development
Chapter 4: Threads Modified by Dr. Neerja Mhaskar for CS 3SH3.
Introduction to threads
Introduction ITEC 420.
Using Ada-C/C++ Changer as a Converter Automatically convert to C/C++ to reuse or redeploy your Ada code Eliminate the need for a costly and.
Chapter 4: Threads.
Chapter 4: Threads.
Advanced Operating Systems CIS 720
MCTS Guide to Microsoft Windows 7
Lighting Up Windows Server 2008 R2 Using the ConcRT on UMS
SQL Server Monitoring Overview
VDK Concepts and Features How to Create a Project with VDK support
runtime verification Brief Overview Grigore Rosu
Many-core Software Development Platforms
Chapter 4: Threads.
Chapter 4: Threads.
Objective Understand the concepts of modern operating systems by investigating the most popular operating system in the current and future market Provide.
Compiler Back End Panel
Compiler Back End Panel
F# for Parallel and Asynchronous Programming
Compiler Front End Panel
LonMaker® Turbo Editions New Features
Software visualization and analysis tool box
CHAPTER 4:THreads Bashair Al-harthi OPERATING SYSTEM
Multithreaded Programming
Chapter 4: Threads & Concurrency
Chapter 4: Threads.
Lecture 2 The Art of Concurrency
Why Threads Are A Bad Idea (for most purposes)
Delivering great hardware solutions for Windows
VDK Concepts and Features How to Create a Project with VDK support
VDK Concepts and Features How to Create a Project with VDK support
HPC User Forum: Back-End Compiler Technology Panel
Intel Parallel Studio Examples
Why Threads Are A Bad Idea (for most purposes)
Chapter 4: Threads.
Objective Understand the concepts of modern operating systems by investigating the most popular operating system in the current and future market Provide.
Why Threads Are A Bad Idea (for most purposes)
.Net for Test and Measurement
Presentation transcript:

Intel® Parallel Studio and Advisor 英特尔软件工具技术支持部 经理 甘驰 V0.5; 22 June 2010

Market Leading Software Tools Optimize Multicore Performance Intel® Parallel Studio If you are using Microsoft Visual Studio* for development and are ready to introduce or extend parallelism in your application, Parallel Studio brings parallelism to your development. Integrated into the Visual Studio* environment, Parallel Studio allows you to immediately be productive as you evaluate your application and add parallelism. Intel® Parallel Composer (C++ compiler, parallel debugger extensions, IPP, TBB) Intel® Parallel Inspector (Memory checker and thread checker) Intel® Parallel Amplifier ( Hotspots, concurrency analysis, locks/waits analysis) Intel® Compilers Intel® C++ Compiler Professional Edition 11.0 for Windows*, Linux, and Mac OS* Intel® Fortran Compiler Professional Edition 11.0 for Windows*, Linux, and Mac OS* Threading performance libraries Intel® Integrated Performance Primitives 6.0 Intel® Math Kernel Library 10.1 Intel® Threading Building Blocks 2.1 Threading performance tools Intel® VTune™ Performance Analyzer 9.1 for Windows and Linux Intel® Thread Checker 3.1 for Windows and Linux Intel® Thread Profiler 3.1 Intel® Cluster Toolkit 3.2^ for Windows* and Linux* Intel® MPI Library 3.2 Intel® MPI Benchmarks 3.2 Intel® Trace Analyzer and Collector 7.2 Intel® Cluster Toolkit Compiler Edition 3.2^ for Windows* and Linux* Intel® C++ Compiler 11.0 Intel® Fortran Compiler 11.0 ^For Linux* or Microsoft Windows* Compute Cluster Server (CCS) 11/15/201811/15/2018

Two Product Lines for Two Needs Focus on Parallel Performance C++ & Fortran* on Windows*, Linux*, Mac OS*X Focus on Parallel Productivity C++ using Visual Studio* on Windows* 11/15/201811/15/2018

Intel® Parallel Studio - revisited all-in-one toolset for the software development lifecycle DESIGN CODE & DEBUG VERIFY TUNE 4

Intel® Parallel Composer CODE & DEBUG PHASE Develop effective applications with a C/C++ compiler and comprehensive threaded libraries CODE & DEBUG PHASE Easier, faster parallelism for Windows* apps Integrated into the Visual Studio* environment C/C++ compiler and advanced threaded libraries Intel® Threading Building Blocks, OpenMP*, Auto-parallelization, Intel® Integrated Performance Primitives Built-in parallel debugger Save time and increase productivity Code and debug phase Combination of compilers, libraries, including OpenMP*, and parallel debugger extension supports easier, faster multi-threading of serial and parallel applications Add parallelism to applications through simple concurrency functions, data parallel arrays, pre-threaded domain-specific libraries, parallel debugging functionality, and more Debugger plug-in simplifies parallel debugging and ensures thread accuracy Library components, including Intel® Threading Building Blocks (Intel® TBB), Intel® Math Kernel Library (Intel® MKL), and Intel® Integrated Performance Primitives (Intel® IPP), provide a variety of threaded generic and application-specific functions 11/15/201811/15/2018

Intel’s Family of Parallel Models New Parallel Models Intel’s Fixed Function Libraries Other Supported Standards Intel® TBB Cilk Intel® MKL OpenMP Intel Ct Technology SIMD Pragma + Array Notations Intel® IPP Co-Array Ftn The earlier foil was overwhelming. It was stated in the presentation and I agree. In an attempt to make it less so, I removed the research content and combined MPI and OpenMP so that we have fewer boxes. It will be good if someone who knows how to use PPT smart art will make the outside frame smaller. MPI Supported by Parallel Composer

Intel® Parallel Inspector VERIFY PHASE Ensure application reliability with proactive memory and threading error checking VERIFY PHASE Find threading errors faster Run from Visual Studio* Parallel memory and threading error checking Single tool to find data races, deadlock and a wide variety of memory errors Rapid analysis of threaded code Help ensure Windows* application reliability Ship apps that run error-free Verify phase Only combined threading and memory checker available today: detects challenging threading and memory errors to help ensure application reliability Fastest and most comprehensive method to pinpoint latent multi-threading errors Rapidly provides root-cause analysis for crash-causing threading and memory defects by monitoring the runtime behavior of the code Maps errors to the source-code line, call stack, and memory reference Helps ensure that shipped applications run error-free on customer systems Identify memory issues in serial and parallel applications in addition to threading errors. 11/15/201811/15/2018

Intel® Parallel Amplifier TUNE PHASE Quickly find bottlenecks and tune serial and parallel applications for scalable multicore performance Quickly find bottlenecks Tune Windows* apps Hotspot Analysis Where does my program spend most of the time? Concurrency Analysis Where and Why doesn’t my program utilize all available core? Tune phase Quickly find multicore performance bottlenecks without needing to know the processor architecture or assembly code Takes away the guesswork and analyzes performance behavior in Windows* applications, providing quick access to scaling information for faster and improved decision making Specifically designed for parallel applications Fine-tuning for optimal performance, ensuring cores are fully exploited and new capabilities are supported Locks&Wait Analysis Where and Why does my program wait? Optimize serial and parallel application performance with 3 easy to use, powerful analysis methods 11/15/201811/15/2018

Intel® Parallel Advisor 11/15/2018

Intel Parallel Advisor Intel Parallel Advisor helps developers identify and experiment with parallel opportunities in existing programs By using Parallel Advisor, the developer can evaluate the return of their investment (performance benefit vs. the cost of transitioning to parallelism) before committing to a course of action…

Intel Parallel Advisor As the user works through the tool, Parallel Advisor analyzes the data collected during the execution of the user’s program: Survey – Helps the developer focus on the hot call trees and loops as locations to experiment with parallelism. The developer will insert Advisor annotations into their sources to describe their parallel experiment. Suitability – Helps developer evaluate the performance of their parallel experiment by displaying the performance projection for each parallel site and how each site’s performance impacts the entire program. Correctness – Helps the developer by identifying data issues (races) of their parallel experiment. Annotation insertion is necessary for Suitability & Correctness tools Annotations are macros that do not impact the users applications when Advisor is disabled/not present – they become “empty statements”. A parallel experiment or parallel modeling is a set of “Annotated tasks” within an “Annotated Site”. Tasks and sites can be in different files, different projects (site in image project, tasks in library). Annotation pairs (site, task, locks) can not span scope BOUNDARIEs

Intel Parallel Advisor – Survey The Survey Report shows the Total & Self Time of call sites & loops Drill down into the Source code Survey – Helps the developer focus on the hot call trees and loops as locations to experiment with parallelism. Top down list of call sites & loops

Intel Parallel Advisor – Suitability Impact to entire program Proposed parallel sites How to maximize the performance benefit of the selected site Annotations that occur within the selected site Suitability – Helps developer evaluate the performance of their parallel experiment by displaying the performance projection for each parallel site and how each site’s performance impacts the entire program. Help when you need it

Intel Parallel Advisor – Correctness Problems and Messages Filters Drilling down into an observation Correctness – Helps the developer by identifying data issues (races) of their parallel experiment. Observation problem sets

Parallel Advisor – Summary Report Displays the developer’s annotated parallel sites & tasks, with their associated Suitability performance projection and Correctness data-sharing problems: This helps the developer identify which parallel experiment to focus on. MySite1 has 6 errors and 3 warnings (Correctness) and self Max Gain of 7.5x – the program could benefit 1.8x

Intel Parallel Advisor Offers guidance through the methodology. Annotation Wizard assists with the creation and insertion of Advisor annotations. Workflow window - The current step is high-lighted.

Intel® Parallel Studio 2011 Family of Parallel Programming Models Intel® Parallel Advisor – parallelism design innovation General Enhancements Supports Visual Studio 2010