Intel(R) Xeon Phi(tm) Advanced Workshop Labs

This is a set of labs we taught during past workshops, intended to cover more advanced concepts. These are written so that you should be able to guide yourself. The labs are available are both in C/C++ and Fortran.

Before you attempt to run these labs, make sure your environment is properly set up.

Lab	Instruction Manuals

MIC

Vectorization

Tuning

MKL

MPI

Intel® Many Integrated Core Architecture

Contrat de licence:

Intel Sample Source Code License Agreement

Improving graphics processing performance using Intel(R) Cilk(TM) Plus

Author(s): Anoop Madhusoodhanan Prabha, Mark Sabahi

Linux*

Microsoft Windows* (XP, Vista, 7)

Microsoft Windows* 8

Apple Mac OS X*

C/C++

Compilateur Intel® C++

Intel® C++ Composer XE

Intel® Cilk™ Plus

Intel® Composer XE

Intel® C++ Studio XE

Intel® Parallel Studio

Intel® Parallel Studio XE

sepia

filter

CilkPlus

Cilk+

Cilk

array notation

cilk_for

AOS

SOA

Vectorization

Requirements for Vectorizable Loops

Vectorization is one of many optimizations that are enabled by default in the latest Intel compilers. In order to be vectorized, loops must obey certain conditions, listed below. Some additional ways to help the compiler to vectorize loops are described.

Linux*

Apple Mac OS X*

Microsoft Windows* (XP, Vista, 7)

C/C++

Fortran

Intermédiaire

Compilateur Intel® C++

Compilateur Intel® Fortran

Intel® Parallel Composer

Vectorization

loops

Optimisation

Being Successful with the Intel® Compilers -- You Need to Know

Tips and techniques on using the Intel Compilers to maximize your application performance.

Linux*

Microsoft Windows* (XP, Vista, 7)

Apple Mac OS X*

C/C++

Fortran

Compilateur Intel® C++

Compilateur Intel® Fortran

Intel® Parallel Composer

Emberson beta program

Vectorization

IPO

floating point

optimize

PGO

interprocedural

profile guided

precision

x87

compiler

Optimisation

SIMD tuning with ASM pt. 4 - Vectorization & ICC

If you remember from my first post I presented a program. Stripping out the setup, the part we care about - the part that does the work - is this:


        float x[PTS], 

        float y[PTS];

        for (int i = 0; i < PTS; i++) { // line 13 in orig source

                x[i] += y[i];           // line 14 in orig source

        }

Vectorization Toolkit

A tool kit that gives 6 Steps to Increase Performance Through Vectorization in Your Application

Linux*

Apple Mac OS X*

Microsoft Windows* (XP, Vista, 7)

C/C++

Fortran

Compilateur Intel® C++

Amplificateur Intel® VTune™ XE

Vectorization

Optimisation

A Guide to Auto-vectorization with Intel® C++ Compilers

How to use the automatic vectorizer of the Intel® C/C++ Compiler to optimize your application using Intel Streaming SIMD Extensions (Intel SSE) or Intel Advanced Vector Extensions (Intel AVX).

Linux*

Microsoft Windows* (XP, Vista, 7)

Apple Mac OS X*

C/C++

SSE

Vectorization

Vectorizer

optimization

optimize

compiler

AVX

Emberson beta program

Optimisation

Minimize frustration and maximize tuning effort with Amdahl's Law

I recently had a question from a customer who had introduced a succesful optimization to a hot function in his application, but did not see as much improvement in the overall application as he expected. This is a fairly common occurence in the iterative process of performance tuning. Usually it happens for one of two reasons.