CAPS Technology. ProHMPT, 2009 March12 th
|
|
- Jayson Brooks
- 5 years ago
- Views:
Transcription
1 CAPS Technology ProHMPT, 2009 March12 th
2 Overview of the Talk 1. HMPP in a nutshell Directives for Hardware Accelerators (HWA) 2. HMPP Code Generation Capabilities Efficient code generation for CUDA 3. Library adapter HPL / DGEMM experiment 4. Codelet Finder
3
4 HMPP Directives C and Fortran directives to program hardware accelerators Ensure portability and default compilation and execution Declare hardware implementations of native functions Indicate resource allocation and communication Place synchronization barriers A standard and portable way of programming A programming glue between general-purpose and hardware-specific languages Insulation of hardware-specific kernels in C and Fortran code HMPP Workbench 4
5 Directives Principles Declare hardware specific implementations of functions (codelets) Can be specialized to the execution context (data size, ) Codelet calls Synchronous, asynchronous properties Data transfers Data preloading Synchronization barriers Host CPU waits until remote computation has completed Main Memory Application data General Purpose Processor Cores Upload Download Remote Procedure call HWA Application data on HWA Cores HMPP Workbench 5
6 Simple Example #pragma hmpp sgemm codelet, target=cuda, args[vout].io=inout extern void sgemm( int m, int n, int k, float alpha, const float vin1[n][n], const float vin2[n][n], float beta, float vout[n][n] ); int main(int argc, char **argv) { for( j = 0 ; j < 2 ; j++ ) { #pragma hmpp sgemm callsite sgemm( size, size, size, alpha, vin1, vin2, beta, vout ); HMPP Workbench 6
7 HMPP Codelet Definition A pure function to be executed in a remote device or specialized core No global variable No side effects Several possible variants for different targets for different use contexts (vector size,...) Managed by the HMPP runtime HMPP API provides the necessary support functions HMPP Workbench 7
8 Directives Overview #pragma hmpp <label> <directive type> [, <directive parameter>]* [&]!$hmpp <label> <directive type> [, <directive parameter>]* [&] A unique label identifies a group of directives that belong to the same codelet Directive types: codelet: codelet declaration callsite: codelet call, can be asynchronous advancedload: preloading of data delegatedstore: wait for data result upload synchronize: wait for the completion of a codelet release: free a compute unit for another codelet HMPP Workbench 8
9 Advanced Programming int main(int argc, char **argv) { #pragma hmpp sgemm allocate, args[vin1;vin2;vout].size={size,size #pragma hmpp sgemm advancedload, args[vin1;vin2;vout] & #pragma hmpp sgemm advancedload, args[m;n;k;alpha;beta] for( j = 0 ; j < 2 ; j++ ) { #pragma hmpp sgemm callsite, asynchronous, & #pragma hmpp sgemm args[vin1;vin2;vout].advancedload=true & #pragma hmpp sgemm args[m;n;k;alpha;beta].advancedload=true sgemm( size, size, size, alpha, vin1, vin2, beta, vout ); #pragma hmpp sgemm synchronize #pragma hmpp sgemm delegatedstore, args[vout] #pragma hmpp sgemm release Allocate and initialize device outside loop Preload data Download result when needed Execute asynchronously HMPP Workbench 9
10 Codelet Directive (1) Declare a hardware specific implementation of a function Several possible variants (target, execution context) Default is the native codelet #pragma hmpp label codelet, target=cuda:brook, args[v1].io=out #pragma hmpp label2 codelet, target=sse, args[v1].io=out, cond= n<800 void MyCodelet(int n, float v1[n], float v2[n], float v3[n]) { int i; for (i = 0 ; i < n ; i++) { v1[i] = v2[i] + v3[i]; HMPP Workbench 10
11 Advancedload Directive (1) Data transfers strongly impact on performance Try to preload data before codelet call site #pragma hmpp simple advancedload, args[v2], asynchronous,\ args[v2].addr= t2 for (k = 0 ; k < iter ; k++) { #pragma hmpp simple callsite, args[v2].advancedload=true simplefunc1(n, &(t1[k*n]), &(t2[k*n]), &(t3[k*n])); #pragma hmpp simple advancedload, args[v2], asynchronous,\ args[v2].addr="&(t2[(k+1)*n]), args[v2].size="(n)" /* do something else */ #pragma hmpp simple release HMPP Workbench 11
12 Advancedload Directive (2) Avoid reloading constant data int main(int argc, char **argv) { #pragma hmpp simple advancedload, args[v2], const for (j=0; j<n; j++){ #pragma hmpp simple callsite, args[v2].advancedload=true simplefunc1(n,t1[j], t2, t3[j], alpha); #pragma hmpp label release t2 is not reloaded each loop iteration HMPP Workbench 12
13 Codelet Generation
14 Objectives Allow to transparently use HWA From C or Fortran to CUDA, Brook, Allow for code tuning at source code level Directives based approach
15 Code Generation Flow
16 Codelet Generation C, Java or Fortran source code input HWA oriented subset of the languages Set of directives to Optimize target codelet generation Express parallelism expression Make code tuning easier Generated code can also be tuned
17 Loop Parallelization Force or prevent the parallelization of loops Help defining kernels in a codelet #pragma hmppcg parallel for (i=0; i < n; i++) { #pragma hmppcg noparallel for (j=0; j < n; j++) { D[i][j] = A[i][j] * E[3][j];
18 Input C Code Example 1 typedef struct{ float r, i; Complex; #pragma hmpp convolution2d codelet, args[data; opx].io=in, args[convr].io=out, target=cuda void convolution2d( Complex *data, int nx, int ny, Complex *opx, int oplx, int oply, Complex *convr) { int hoplx = (oplx+1)/2; int hoply = (oply+1)/2; int iy, ix; #pragma hmppcg parallel for (iy = 0; iy < ny; iy++) { #pragma hmppcg parallel for (ix = 0; ix < nx; ix++) { float dumr =0.0, dumi = 0.0; int ky; for(ky = -(oply - hoply - 1); ky <= hoply; ky++) { int kx; for(kx = -(oplx - hoplx - 1); kx <= hoplx; kx++){ int dx = min( max(ix+kx, 0), (nx - 1) ); int dy = min( max(iy+ky, 0), (ny - 1) ); dumr += data[dy * nx + dx].r * opx[(hoply - ky) * oplx + (hoplx - kx)].r; dumr -= data[dy * nx + dx].i * opx[(hoply - ky) * oplx + (hoplx - kx)].i; dumi += data[dy * nx + dx].r * opx[(hoply - ky) * oplx + (hoplx - kx)].i; dumi += data[dy * nx + dx].i * opx[(hoply - ky) * oplx + (hoplx - kx)].r; convr[iy*nx+ix].r = dumr; convr[iy*nx+ix].i = dumi;
19 Input Fortran Code Example 2!$HMPP sgemm3 codelet, target=cuda, args[vout].io=inout SUBROUTINE sgemm(m,n,k2,alpha,vin1,vin2,beta,vout) INTEGER, INTENT(IN) :: m,n,k2 REAL, INTENT(IN) :: alpha,beta REAL, INTENT(IN) :: vin1(n,n), vin2(n,n) REAL, INTENT(INOUT) :: vout(n,n) REAL :: prod INTEGER :: i,j,k!$hmppcg unroll(8), jam(2), noremainder!$hmppcg parallel DO j=1,n!$hmppcg unroll(8), splitted, noremainder!$hmppcg parallel DO i=1,n prod = 0.0 DO k=1,n prod = prod + vin1(i,k) * vin2(k,j) ENDDO vout(i,j) = alpha * prod + beta * vout(i,j) ; END DO END DO END SUBROUTINE sgemm
20 MxM Performance
21 Performance Examples
22 Tuning Issue Example #pragma hmpp astex_codelet 1 codelet & #pragma hmpp astex_codelet 1, args[c].io=in & #pragma hmpp astex_codelet 1, args[v].io=inout & #pragma hmpp astex_codelet 1, args[u].io=inout & #pragma hmpp astex_codelet 1, target=cuda & #pragma hmpp astex_codelet 1, version=1.4.0 void astex_codelet 1(float u[256][256][256], float v[256][256][256], float c[256][256][256], const int K, const float x2){ astex_thread_begin:{ for (int it = 0 ; it < K ; ++it){ for (int i2 = 1 ; i2 < ; ++i2){ Need interchange for (int i3 = 1 ; i3 < ; ++i3){ for (int i1 = 1 ; i1 < ; ++i1){ float coeff = c[i3][i2][i1] * c[i3][i2][i1] * x2; float sum = u[i3][i2][i1 + 1] + u[i3][i2][i1-1]; sum += u[i3][i2 + 1][i1] + u[i3][i2-1][i1]; sum += u[i3 + 1][i2][i1] + u[i3-1][i2][i1]; v[i3][i2][i1] = ( * coeff) * u[i3][i2][i1] + coeff * sum - v[i3][i2][i1]; for (int i2 = 1 ; i2 < ; ++i2){ for (int i3 = 1 ; i3 < ; ++i3){ for (int i1 = 1 ; i1 < ; ++i1{..... astex_thread_end:;
23 Library Issues
24 Motivations Various implementations of libraries are available for a given target CUBLAS, MKL, ATLAS, No strict performance order Each library has a different performance profile Best choice depends on platform and runtime parameters User left with a complex issue Routine performance measure Decision programming Hardware version adaptation Development partially funded by STREP Milepost Machine Learning for Embedded Programs Optimisation
25 Difficult Decisions Making with Alternative Codes (Multiversioning) Various implementations of routines are available or can be generated for a given target CUBLAS, MKL, ATLAS, SIMD instructions, GPcore, HWA, Hybrid No strict performance order Each implementation has a different performance profile Best choice depends on platform and runtime parameters Decision is a complex issue How to produce the decision?
26 Library Adapter Overview
27 Illustrating Example: Dealing with Multiple BLAS Implementations Runtime selection of DGEMM in High Performance Linpack Intel(R) Xeon(R) 2.50GHz CUBLAS - Tesla C1060, Intel MKL Three binaries of the application Static linking with CUBLAS Static linking with MKL Library mix with selection of routine at runtime Automatically generated using CAPS tooling Three hardware resource configurations GPU + 1, 2, and 4 cores used for MKL
28 Performance Using One Core Performance in Gigaflops 4 problem sizes: 64, 500, 1200, ,3 20 Performance (GFLOPS) ,3 6,5 8 8,1 9 Cublas MKL Dyn. Sel ,4 1,3 1,4 1,2 0, Problem Size
29 Performance Using Two Cores Performance (GFLOPS) Cublas MKL Dyn. Sel. 10 6,5 7,6 5 4,3 4,4 0 0,07 1,4 1,2 0, Problem Size
30 Performance Using Four Cores Performance (GFLOPS) ,2 9, Cublas MKL Dyn. Sel ,4 0 0,07 0,9 1,2 1, Problem Size
31 Codelet Finder Alpha version
32 Codelet Finder Overview Partitioning of C code to highlight codelets Data value specialization Aliasing speculation static static Useful for dynamic dynamic HWA exploitation (and maybe vectorization and parallelization) Partitioned code
33 Extracted Codelets Are Not Just Hotspots HWA data mapping in local memory adds constraints 0xA 10 { for (x = 0 ; x < i_size ; x++) { diff[x + y * i_size] = pix1[x] - pix2[x]; pix1 += i_pix1; pix2 += i_pix2; Main memory 0x0 05 HWA local memory 0xA 40 pix1= 0xA 10 0x0 35 pix1= 0x0 05
34 Example of Partitioning to Use HWA (1) #pragma hmpp astex_codelet 1 codelet & #pragma hmpp astex_codelet 1, args[c].io=in & #pragma hmpp astex_codelet 1, args[v].io=inout & #pragma hmpp astex_codelet 1, args[u].io=inout & #pragma hmpp astex_codelet 1, target=cuda & #pragma hmpp astex_codelet 1, version=1.4.0 void astex_codelet 1(float u[256][256][256], float v[256][256][256], float c[256][256][256], const int K, const float x2){ astex_thread_begin:{ for (int it = 0 ; it < K ; ++it){ for (int i2 = 1 ; i2 < ; ++i2){ for (int i3 = 1 ; i3 < ; ++i3){ for (int i1 = 1 ; i1 < ; ++i1){ float coeff = c[i3][i2][i1] * c[i3][i2][i1] * x2; float sum = u[i3][i2][i1 + 1] + u[i3][i2][i1-1]; sum += u[i3][i2 + 1][i1] + u[i3][i2-1][i1]; sum += u[i3 + 1][i2][i1] + u[i3-1][i2][i1]; v[i3][i2][i1] = ( * coeff) * u[i3][i2][i1] + coeff * sum - v[i3][i2][i1]; for (int i2 = 1 ; i2 < ; ++i2){ for (int i3 = 1 ; i3 < ; ++i3){ for (int i1 = 1 ; i1 < ; ++i1{..... astex_thread_end:;
35 Example of Partitioning to Use HWA (2) Extract codelet to be executed on HWA Data specialization Aliasing speculation Convolution code icc -O3 vs HWA Speedup is 3.4 with the HWA Codelet tuning Loop interchange was needed
36 Codelet Finder in ProHMPT Can be used to provide codelet testbed for the various techniques
37 Conclusion
38 CAPS in ProHMPT Génération de codes adaptatifs Définition de directives Allocation dynamiques de ressources
39 Tasks Thème 2: extraction du parallélisme Tâche 3: Compilation et analyse du code statique Fondé sur DPIL Tâche 4: Langage Extension d OpenMP pour l hétérogène Thème 3: Support logiciel Tâche 7: Ordonnancement Allocation de ressources
Parallel Hybrid Computing Stéphane Bihan, CAPS
Parallel Hybrid Computing Stéphane Bihan, CAPS Introduction Main stream applications will rely on new multicore / manycore architectures It is about performance not parallelism Various heterogeneous hardware
More informationDealing with Heterogeneous Multicores
Dealing with Heterogeneous Multicores François Bodin INRIA-UIUC, June 12 th, 2009 Introduction Main stream applications will rely on new multicore / manycore architectures It is about performance not parallelism
More informationParallel Hybrid Computing F. Bodin, CAPS Entreprise
Parallel Hybrid Computing F. Bodin, CAPS Entreprise Introduction Main stream applications will rely on new multicore / manycore architectures It is about performance not parallelism Various heterogeneous
More informationHeterogeneous multicore parallel programming for graphics processing units
Scientific Programming 17 (2009) 325 335 325 DOI 10.3233/SPR-2009-0292 IOS Press Heterogeneous multicore parallel programming for graphics processing units Francois Bodin and Stephane Bihan CAPS entreprise,
More informationHeterogeneous Multicore Parallel Programming
Innovative software for manycore paradigms Heterogeneous Multicore Parallel Programming S. Chauveau & L. Morin & F. Bodin Introduction Numerous legacy applications can benefit from GPU computing Many programming
More informationParallel Hybrid Computing F. Bodin, CAPS Entreprise
Parallel Hybrid Computing F. Bodin, CAPS Entreprise Introduction Main stream applications will rely on new multicore / manycore architectures It is about performance not parallelism Various heterogeneous
More informationHMPP port. G. Colin de Verdière (CEA)
HMPP port G. Colin de Verdière (CEA) Overview.Uchu prototype HMPP MOD2AS MOD2AM HMPP in a real code 2 The UCHU prototype Bull servers 1 login node 4 nodes 2 Haperton, 8GB 2 NVIDIA Tesla S1070 IB DDR Slurm
More informationMIGRATION OF LEGACY APPLICATIONS TO HETEROGENEOUS ARCHITECTURES Francois Bodin, CTO, CAPS Entreprise. June 2011
MIGRATION OF LEGACY APPLICATIONS TO HETEROGENEOUS ARCHITECTURES Francois Bodin, CTO, CAPS Entreprise June 2011 FREE LUNCH IS OVER, CODES HAVE TO MIGRATE! Many existing legacy codes needs to migrate to
More informationPragma-based GPU Programming and HMPP Workbench. Scott Grauer-Gray
Pragma-based GPU Programming and HMPP Workbench Scott Grauer-Gray Pragma-based GPU programming Write programs for GPU processing without (directly) using CUDA/OpenCL Place pragmas to drive processing on
More informationIncremental Migration of C and Fortran Applications to GPGPU using HMPP HPC Advisory Council China Conference 2010
Innovative software for manycore paradigms Incremental Migration of C and Fortran Applications to GPGPU using HMPP HPC Advisory Council China Conference 2010 Introduction Many applications can benefit
More informationAddressing Heterogeneity in Manycore Applications
Addressing Heterogeneity in Manycore Applications RTM Simulation Use Case stephane.bihan@caps-entreprise.com Oil&Gas HPC Workshop Rice University, Houston, March 2008 www.caps-entreprise.com Introduction
More informationCode Migration Methodology for Heterogeneous Systems
Code Migration Methodology for Heterogeneous Systems Directives based approach using HMPP - OpenAcc F. Bodin, CAPS CTO Introduction Many-core computing power comes from parallelism o Multiple forms of
More informationJCudaMP: OpenMP/Java on CUDA
JCudaMP: OpenMP/Java on CUDA Georg Dotzler, Ronald Veldema, Michael Klemm Programming Systems Group Martensstraße 3 91058 Erlangen Motivation Write once, run anywhere - Java Slogan created by Sun Microsystems
More informationOpenACC Standard. Credits 19/07/ OpenACC, Directives for Accelerators, Nvidia Slideware
OpenACC Standard Directives for Accelerators Credits http://www.openacc.org/ o V1.0: November 2011 Specification OpenACC, Directives for Accelerators, Nvidia Slideware CAPS OpenACC Compiler, HMPP Workbench
More informationLocality-Aware Automatic Parallelization for GPGPU with OpenHMPP Directives
Locality-Aware Automatic Parallelization for GPGPU with OpenHMPP Directives José M. Andión, Manuel Arenaz, François Bodin, Gabriel Rodríguez and Juan Touriño 7th International Symposium on High-Level Parallel
More informationAcceleration of a Python-based Tsunami Modelling Application via CUDA and OpenHMPP
Acceleration of a Python-based Tsunami Modelling Application via CUDA and OpenHMPP Zhe Weng and Peter Strazdins*, Computer Systems Group, Research School of Computer Science, The Australian National University
More informationSolving Dense Linear Systems on Graphics Processors
Solving Dense Linear Systems on Graphics Processors Sergio Barrachina Maribel Castillo Francisco Igual Rafael Mayo Enrique S. Quintana-Ortí High Performance Computing & Architectures Group Universidad
More informationA Uniform Programming Model for Petascale Computing
A Uniform Programming Model for Petascale Computing Barbara Chapman University of Houston WPSE 2009, Tsukuba March 25, 2009 High Performance Computing and Tools Group http://www.cs.uh.edu/~hpctools Agenda
More informationHow to Write Code that Will Survive the Many-Core Revolution
How to Write Code that Will Survive the Many-Core Revolution Write Once, Deploy Many(-Cores) Guillaume BARAT, EMEA Sales Manager CAPS worldwide ecosystem Customers Business Partners Involved in many European
More informationAn Hybrid Data Transfer Optimization Technique for GPGPU
An Hybrid Data Transfer Optimization Technique for GPGPU Eric Petit, François Bodin, Romain Dolbeau To cite this version: Eric Petit, François Bodin, Romain Dolbeau. An Hybrid Data Transfer Optimization
More informationHow to write code that will survive the many-core revolution Write once, deploy many(-cores) F. Bodin, CTO
How to write code that will survive the many-core revolution Write once, deploy many(-cores) F. Bodin, CTO Foreword How to write code that will survive the many-core revolution? is being setup as a collective
More informationPerformance Analysis of Memory Transfers and GEMM Subroutines on NVIDIA TESLA GPU Cluster
Performance Analysis of Memory Transfers and GEMM Subroutines on NVIDIA TESLA GPU Cluster Veerendra Allada, Troy Benjegerdes Electrical and Computer Engineering, Ames Laboratory Iowa State University &
More informationCUDA 6.0 Performance Report. April 2014
CUDA 6. Performance Report April 214 1 CUDA 6 Performance Report CUDART CUDA Runtime Library cufft Fast Fourier Transforms Library cublas Complete BLAS Library cusparse Sparse Matrix Library curand Random
More informationLittle Motivation Outline Introduction OpenMP Architecture Working with OpenMP Future of OpenMP End. OpenMP. Amasis Brauch German University in Cairo
OpenMP Amasis Brauch German University in Cairo May 4, 2010 Simple Algorithm 1 void i n c r e m e n t e r ( short a r r a y ) 2 { 3 long i ; 4 5 for ( i = 0 ; i < 1000000; i ++) 6 { 7 a r r a y [ i ]++;
More informationNVIDIA GTX200: TeraFLOPS Visual Computing. August 26, 2008 John Tynefield
NVIDIA GTX200: TeraFLOPS Visual Computing August 26, 2008 John Tynefield 2 Outline Execution Model Architecture Demo 3 Execution Model 4 Software Architecture Applications DX10 OpenGL OpenCL CUDA C Host
More informationAccelerated Test Execution Using GPUs
Accelerated Test Execution Using GPUs Vanya Yaneva Supervisors: Ajitha Rajan, Christophe Dubach Mathworks May 27, 2016 The Problem Software testing is time consuming Functional testing The Problem Software
More informationOpenACC Course. Office Hour #2 Q&A
OpenACC Course Office Hour #2 Q&A Q1: How many threads does each GPU core have? A: GPU cores execute arithmetic instructions. Each core can execute one single precision floating point instruction per cycle
More informationOverview of research activities Toward portability of performance
Overview of research activities Toward portability of performance Do dynamically what can t be done statically Understand evolution of architectures Enable new programming models Put intelligence into
More informationAn Extension of the StarSs Programming Model for Platforms with Multiple GPUs
An Extension of the StarSs Programming Model for Platforms with Multiple GPUs Eduard Ayguadé 2 Rosa M. Badia 2 Francisco Igual 1 Jesús Labarta 2 Rafael Mayo 1 Enrique S. Quintana-Ortí 1 1 Departamento
More informationPROGRAMOVÁNÍ V C++ CVIČENÍ. Michal Brabec
PROGRAMOVÁNÍ V C++ CVIČENÍ Michal Brabec PARALLELISM CATEGORIES CPU? SSE Multiprocessor SIMT - GPU 2 / 17 PARALLELISM V C++ Weak support in the language itself, powerful libraries Many different parallelization
More informationAccelerating Financial Applications on the GPU
Accelerating Financial Applications on the GPU Scott Grauer-Gray Robert Searles William Killian John Cavazos Department of Computer and Information Science University of Delaware Sixth Workshop on General
More informationParallel Programming. Libraries and Implementations
Parallel Programming Libraries and Implementations Reusing this material This work is licensed under a Creative Commons Attribution- NonCommercial-ShareAlike 4.0 International License. http://creativecommons.org/licenses/by-nc-sa/4.0/deed.en_us
More informationSolving Dense Linear Systems on Platforms with Multiple Hardware Accelerators
Solving Dense Linear Systems on Platforms with Multiple Hardware Accelerators Francisco D. Igual Enrique S. Quintana-Ortí Gregorio Quintana-Ortí Universidad Jaime I de Castellón (Spain) Robert A. van de
More informationParallel Programming Libraries and implementations
Parallel Programming Libraries and implementations Partners Funding Reusing this material This work is licensed under a Creative Commons Attribution- NonCommercial-ShareAlike 4.0 International License.
More informationAuto-tuning a High-Level Language Targeted to GPU Codes. By Scott Grauer-Gray, Lifan Xu, Robert Searles, Sudhee Ayalasomayajula, John Cavazos
Auto-tuning a High-Level Language Targeted to GPU Codes By Scott Grauer-Gray, Lifan Xu, Robert Searles, Sudhee Ayalasomayajula, John Cavazos GPU Computing Utilization of GPU gives speedup on many algorithms
More informationFahad Zafar, Dibyajyoti Ghosh, Lawrence Sebald, Shujia Zhou. University of Maryland Baltimore County
Accelerating a climate physics model with OpenCL Fahad Zafar, Dibyajyoti Ghosh, Lawrence Sebald, Shujia Zhou University of Maryland Baltimore County Introduction The demand to increase forecast predictability
More informationPorting Scientific Research Codes to GPUs with CUDA Fortran: Incompressible Fluid Dynamics using the Immersed Boundary Method
Porting Scientific Research Codes to GPUs with CUDA Fortran: Incompressible Fluid Dynamics using the Immersed Boundary Method Josh Romero, Massimiliano Fatica - NVIDIA Vamsi Spandan, Roberto Verzicco -
More informationPortability of OpenMP Offload Directives Jeff Larkin, OpenMP Booth Talk SC17
Portability of OpenMP Offload Directives Jeff Larkin, OpenMP Booth Talk SC17 11/27/2017 Background Many developers choose OpenMP in hopes of having a single source code that runs effectively anywhere (performance
More informationMAGMA a New Generation of Linear Algebra Libraries for GPU and Multicore Architectures
MAGMA a New Generation of Linear Algebra Libraries for GPU and Multicore Architectures Stan Tomov Innovative Computing Laboratory University of Tennessee, Knoxville OLCF Seminar Series, ORNL June 16, 2010
More informationC6000 Compiler Roadmap
C6000 Compiler Roadmap CGT v7.4 CGT v7.3 CGT v7. CGT v8.0 CGT C6x v8. CGT Longer Term In Development Production Early Adopter Future CGT v7.2 reactive Current 3H2 4H 4H2 H H2 Future CGT C6x v7.3 Control
More informationOpenACC Fundamentals. Steve Abbott November 15, 2017
OpenACC Fundamentals Steve Abbott , November 15, 2017 AGENDA Data Regions Deep Copy 2 while ( err > tol && iter < iter_max ) { err=0.0; JACOBI ITERATION #pragma acc parallel loop reduction(max:err)
More informationOpenACC. Part I. Ned Nedialkov. McMaster University Canada. October 2016
OpenACC. Part I Ned Nedialkov McMaster University Canada October 2016 Outline Introduction Execution model Memory model Compiling pgaccelinfo Example Speedups Profiling c 2016 Ned Nedialkov 2/23 Why accelerators
More informationINTRODUCTION TO OPENACC. Analyzing and Parallelizing with OpenACC, Feb 22, 2017
INTRODUCTION TO OPENACC Analyzing and Parallelizing with OpenACC, Feb 22, 2017 Objective: Enable you to to accelerate your applications with OpenACC. 2 Today s Objectives Understand what OpenACC is and
More informationCUDA Accelerated Compute Libraries. M. Naumov
CUDA Accelerated Compute Libraries M. Naumov Outline Motivation Why should you use libraries? CUDA Toolkit Libraries Overview of performance CUDA Proprietary Libraries Address specific markets Third Party
More informationIntroduction to CUDA
Introduction to CUDA Overview HW computational power Graphics API vs. CUDA CUDA glossary Memory model, HW implementation, execution Performance guidelines CUDA compiler C/C++ Language extensions Limitations
More informationUsing JURECA's GPU Nodes
Mitglied der Helmholtz-Gemeinschaft Using JURECA's GPU Nodes Willi Homberg Supercomputing Centre Jülich (JSC) Introduction to the usage and programming of supercomputer resources in Jülich 23-24 May 2016
More informationAn Introduction to OpenAcc
An Introduction to OpenAcc ECS 158 Final Project Robert Gonzales Matthew Martin Nile Mittow Ryan Rasmuss Spring 2016 1 Introduction: What is OpenAcc? OpenAcc stands for Open Accelerators. Developed by
More informationOpenACC (Open Accelerators - Introduced in 2012)
OpenACC (Open Accelerators - Introduced in 2012) Open, portable standard for parallel computing (Cray, CAPS, Nvidia and PGI); introduced in 2012; GNU has an incomplete implementation. Uses directives in
More informationHigh Level Programming for GPGPU. Jason Yang Justin Hensley
Jason Yang Justin Hensley Outline Brook+ Brook+ demonstration on R670 AMD IL 2 Brook+ Introduction 3 What is Brook+? Brook is an extension to the C-language for stream programming originally developed
More informationHow to perform HPL on CPU&GPU clusters. Dr.sc. Draško Tomić
How to perform HPL on CPU&GPU clusters Dr.sc. Draško Tomić email: drasko.tomic@hp.com Forecasting is not so easy, HPL benchmarking could be even more difficult Agenda TOP500 GPU trends Some basics about
More informationAn Extension of XcalableMP PGAS Lanaguage for Multi-node GPU Clusters
An Extension of XcalableMP PGAS Lanaguage for Multi-node Clusters Jinpil Lee, Minh Tuan Tran, Tetsuya Odajima, Taisuke Boku and Mitsuhisa Sato University of Tsukuba 1 Presentation Overview l Introduction
More informationNVIDIA Think about Computing as Heterogeneous One Leo Liao, 1/29/2106, NTU
NVIDIA Think about Computing as Heterogeneous One Leo Liao, 1/29/2106, NTU GPGPU opens the door for co-design HPC, moreover middleware-support embedded system designs to harness the power of GPUaccelerated
More informationParallel Programming. Exploring local computational resources OpenMP Parallel programming for multiprocessors for loops
Parallel Programming Exploring local computational resources OpenMP Parallel programming for multiprocessors for loops Single computers nowadays Several CPUs (cores) 4 to 8 cores on a single chip Hyper-threading
More informationAdrian Tate XK6 / openacc workshop Manno, Mar
Adrian Tate XK6 / openacc workshop Manno, Mar6-7 2012 1 Overview & Philosophy Two modes of usage Contents Present contents Upcoming releases Optimization of libsci_acc Autotuning Adaptation Asynchronous
More informationAutotuning and Specialization: Speeding up Matrix Multiply for Small Matrices with Compiler Technology
Autotuning and Specialization: Speeding up Matrix Multiply for Small Matrices with Compiler Technology Jaewook Shin 1, Mary W. Hall 2, Jacqueline Chame 3, Chun Chen 2, Paul D. Hovland 1 1 ANL/MCS 2 University
More informationProfiling and Parallelizing with the OpenACC Toolkit OpenACC Course: Lecture 2 October 15, 2015
Profiling and Parallelizing with the OpenACC Toolkit OpenACC Course: Lecture 2 October 15, 2015 Oct 1: Introduction to OpenACC Oct 6: Office Hours Oct 15: Profiling and Parallelizing with the OpenACC Toolkit
More informationUsing OpenACC With CUDA Libraries
Using OpenACC With CUDA Libraries John Urbanic with NVIDIA Pittsburgh Supercomputing Center Copyright 2015 3 Ways to Accelerate Applications Applications Libraries Drop-in Acceleration CUDA Libraries are
More informationECE 574 Cluster Computing Lecture 10
ECE 574 Cluster Computing Lecture 10 Vince Weaver http://www.eece.maine.edu/~vweaver vincent.weaver@maine.edu 1 October 2015 Announcements Homework #4 will be posted eventually 1 HW#4 Notes How granular
More informationS Comparing OpenACC 2.5 and OpenMP 4.5
April 4-7, 2016 Silicon Valley S6410 - Comparing OpenACC 2.5 and OpenMP 4.5 James Beyer, NVIDIA Jeff Larkin, NVIDIA GTC16 April 7, 2016 History of OpenMP & OpenACC AGENDA Philosophical Differences Technical
More informationProgramming paradigms for GPU devices
Programming paradigms for GPU devices OpenAcc Introduction Sergio Orlandini s.orlandini@cineca.it 1 OpenACC introduction express parallelism optimize data movements practical examples 2 3 Ways to Accelerate
More informationOmpSs + OpenACC Multi-target Task-Based Programming Model Exploiting OpenACC GPU Kernel
www.bsc.es OmpSs + OpenACC Multi-target Task-Based Programming Model Exploiting OpenACC GPU Kernel Guray Ozen guray.ozen@bsc.es Exascale in BSC Marenostrum 4 (13.7 Petaflops ) General purpose cluster (3400
More informationOpenMP programming. Thomas Hauser Director Research Computing Research CU-Boulder
OpenMP programming Thomas Hauser Director Research Computing thomas.hauser@colorado.edu CU meetup 1 Outline OpenMP Shared-memory model Parallel for loops Declaring private variables Critical sections Reductions
More informationSPOC : GPGPU programming through Stream Processing with OCaml
SPOC : GPGPU programming through Stream Processing with OCaml Mathias Bourgoin - Emmanuel Chailloux - Jean-Luc Lamotte January 23rd, 2012 GPGPU Programming Two main frameworks Cuda OpenCL Different Languages
More informationAsian Option Pricing on cluster of GPUs: First Results
Asian Option Pricing on cluster of GPUs: First Results (ANR project «GCPMF») S. Vialle SUPELEC L. Abbas-Turki ENPC With the help of P. Mercier (SUPELEC). Previous work of G. Noaje March-June 2008. 1 Building
More informationCOMP4510 Introduction to Parallel Computation. Shared Memory and OpenMP. Outline (cont d) Shared Memory and OpenMP
COMP4510 Introduction to Parallel Computation Shared Memory and OpenMP Thanks to Jon Aronsson (UofM HPC consultant) for some of the material in these notes. Outline (cont d) Shared Memory and OpenMP Including
More informationComparing OpenACC 2.5 and OpenMP 4.1 James C Beyer PhD, Sept 29 th 2015
Comparing OpenACC 2.5 and OpenMP 4.1 James C Beyer PhD, Sept 29 th 2015 Abstract As both an OpenMP and OpenACC insider I will present my opinion of the current status of these two directive sets for programming
More informationMPI_Send(a,..., MPI_COMM_WORLD); MPI_Recv(a,..., MPI_COMM_WORLD, &status);
$ $ 2 global void kernel(int a[max], int llimit, int ulimit) {... } : int main(int argc, char *argv[]){ MPI_Int(&argc, &argc); MPI_Comm_rank(MPI_COMM_WORLD, &rank); MPI_Comm_size(MPI_COMM_WORLD, &size);
More informationPathScale ENZO GTC12 S0631 Programming Heterogeneous Many-Cores Using Directives. C. Bergström May 14th, 2012
PathScale ENZO GTC12 S0631 Programming Heterogeneous Many-Cores Using Directives C. Bergström May 14th, 2012 Brief Introduction to ENZO 2 PathScale GTC12 S0631 Tutorial May 14th, 2012 ENZO Overview & Goals
More informationIntroduction to the Xeon Phi programming model. Fabio AFFINITO, CINECA
Introduction to the Xeon Phi programming model Fabio AFFINITO, CINECA What is a Xeon Phi? MIC = Many Integrated Core architecture by Intel Other names: KNF, KNC, Xeon Phi... Not a CPU (but somewhat similar
More informationOpenMP API Version 5.0
OpenMP API Version 5.0 (or: Pretty Cool & New OpenMP Stuff) Michael Klemm Chief Executive Officer OpenMP Architecture Review Board michael.klemm@openmp.org Architecture Review Board The mission of the
More informationCUDA Toolkit 4.0 Performance Report. June, 2011
CUDA Toolkit 4. Performance Report June, 211 CUDA Math Libraries High performance math routines for your applications: cufft Fast Fourier Transforms Library cublas Complete BLAS Library cusparse Sparse
More informationOMP2HMPP: HMPP Source Code Generation from Programs with Pragma Extensions
OMP2HMPP: HMPP Source Code Generation from Programs with Pragma Extensions Albert Saà-Garriga Universitat Autonòma de Barcelona Edifici Q,Campus de la UAB Bellaterra, Spain albert.saa@uab.cat David Castells-Rufas
More informationCafeGPI. Single-Sided Communication for Scalable Deep Learning
CafeGPI Single-Sided Communication for Scalable Deep Learning Janis Keuper itwm.fraunhofer.de/ml Competence Center High Performance Computing Fraunhofer ITWM, Kaiserslautern, Germany Deep Neural Networks
More informationTechnische Universität München. GPU Programming. Rüdiger Westermann Chair for Computer Graphics & Visualization. Faculty of Informatics
GPU Programming Rüdiger Westermann Chair for Computer Graphics & Visualization Faculty of Informatics Overview Programming interfaces and support libraries The CUDA programming abstraction An in-depth
More informationParallel Programming in C with MPI and OpenMP
Parallel Programming in C with MPI and OpenMP Michael J. Quinn Chapter 17 Shared-memory Programming 1 Outline n OpenMP n Shared-memory model n Parallel for loops n Declaring private variables n Critical
More informationDistributed Dense Linear Algebra on Heterogeneous Architectures. George Bosilca
Distributed Dense Linear Algebra on Heterogeneous Architectures George Bosilca bosilca@eecs.utk.edu Centraro, Italy June 2010 Factors that Necessitate to Redesign of Our Software» Steepness of the ascent
More informationOvercoming the Barriers to Sustained Petaflop Performance
Overcoming the Barriers to Sustained Petaflop Performance William D. Gropp Mathematics and Computer Science www.mcs.anl.gov/~gropp Argonne National Laboratory A Laboratory Operated by The University of
More informationIntel Array Building Blocks (Intel ArBB) Technical Presentation
Intel Array Building Blocks (Intel ArBB) Technical Presentation Copyright 2010, Intel Corporation. All rights reserved. 1 Noah Clemons Software And Services Group Developer Products Division Performance
More informationAutomatic Performance Tuning. Jeremy Johnson Dept. of Computer Science Drexel University
Automatic Performance Tuning Jeremy Johnson Dept. of Computer Science Drexel University Outline Scientific Computation Kernels Matrix Multiplication Fast Fourier Transform (FFT) Automated Performance Tuning
More informationMPI and OpenMP (Lecture 25, cs262a) Ion Stoica, UC Berkeley November 19, 2016
MPI and OpenMP (Lecture 25, cs262a) Ion Stoica, UC Berkeley November 19, 2016 Message passing vs. Shared memory Client Client Client Client send(msg) recv(msg) send(msg) recv(msg) MSG MSG MSG IPC Shared
More informationOmpSs Fundamentals. ISC 2017: OpenSuCo. Xavier Teruel
OmpSs Fundamentals ISC 2017: OpenSuCo Xavier Teruel Outline OmpSs brief introduction OmpSs overview and influence in OpenMP Execution model and parallelization approaches Memory model and target copies
More informationOpenACC. Introduction and Evolutions Sebastien Deldon, GPU Compiler engineer
OpenACC Introduction and Evolutions Sebastien Deldon, GPU Compiler engineer 3 WAYS TO ACCELERATE APPLICATIONS Applications Libraries Compiler Directives Programming Languages Easy to use Most Performance
More informationThe CLAW project. Valentin Clément, Xavier Lapillonne. CLAW provides high-level Abstractions for Weather and climate models
Federal Department of Home Affairs FDHA Federal Office of Meteorology and Climatology MeteoSwiss The CLAW project CLAW provides high-level Abstractions for Weather and climate models Valentin Clément,
More informationParallel Computing. November 20, W.Homberg
Mitglied der Helmholtz-Gemeinschaft Parallel Computing November 20, 2017 W.Homberg Why go parallel? Problem too large for single node Job requires more memory Shorter time to solution essential Better
More informationOpenACC2 vs.openmp4. James Lin 1,2 and Satoshi Matsuoka 2
2014@San Jose Shanghai Jiao Tong University Tokyo Institute of Technology OpenACC2 vs.openmp4 he Strong, the Weak, and the Missing to Develop Performance Portable Applica>ons on GPU and Xeon Phi James
More informationCSCI 402: Computer Architectures. Parallel Processors (2) Fengguang Song Department of Computer & Information Science IUPUI.
CSCI 402: Computer Architectures Parallel Processors (2) Fengguang Song Department of Computer & Information Science IUPUI 6.6 - End Today s Contents GPU Cluster and its network topology The Roofline performance
More informationCompiling for GPUs. Adarsh Yoga Madhav Ramesh
Compiling for GPUs Adarsh Yoga Madhav Ramesh Agenda Introduction to GPUs Compute Unified Device Architecture (CUDA) Control Structure Optimization Technique for GPGPU Compiler Framework for Automatic Translation
More informationOpen Multi-Processing: Basic Course
HPC2N, UmeåUniversity, 901 87, Sweden. May 26, 2015 Table of contents Overview of Paralellism 1 Overview of Paralellism Parallelism Importance Partitioning Data Distributed Memory Working on Abisko 2 Pragmas/Sentinels
More informationOpenACC 2.5 and Beyond. Michael Wolfe PGI compiler engineer
OpenACC 2.5 and Beyond Michael Wolfe PGI compiler engineer michael.wolfe@pgroup.com OpenACC Timeline 2008 PGI Accelerator Model (targeting NVIDIA GPUs) 2011 OpenACC 1.0 (targeting NVIDIA GPUs, AMD GPUs)
More informationGPU GPU CPU. Raymond Namyst 3 Samuel Thibault 3 Olivier Aumage 3
/CPU,a),2,2 2,2 Raymond Namyst 3 Samuel Thibault 3 Olivier Aumage 3 XMP XMP-dev CPU XMP-dev/StarPU XMP-dev XMP CPU StarPU CPU /CPU XMP-dev/StarPU N /CPU CPU. Graphics Processing Unit GP General-Purpose
More informationTechnology for a better society. hetcomp.com
Technology for a better society hetcomp.com 1 J. Seland, C. Dyken, T. R. Hagen, A. R. Brodtkorb, J. Hjelmervik,E Bjønnes GPU Computing USIT Course Week 16th November 2011 hetcomp.com 2 9:30 10:15 Introduction
More informationGPU Programming Using NVIDIA CUDA
GPU Programming Using NVIDIA CUDA Siddhante Nangla 1, Professor Chetna Achar 2 1, 2 MET s Institute of Computer Science, Bandra Mumbai University Abstract: GPGPU or General-Purpose Computing on Graphics
More informationAdvanced OpenMP Features
Christian Terboven, Dirk Schmidl IT Center, RWTH Aachen University Member of the HPC Group {terboven,schmidl@itc.rwth-aachen.de IT Center der RWTH Aachen University Vectorization 2 Vectorization SIMD =
More informationOpenMP: Open Multiprocessing
OpenMP: Open Multiprocessing Erik Schnetter June 7, 2012, IHPC 2012, Iowa City Outline 1. Basic concepts, hardware architectures 2. OpenMP Programming 3. How to parallelise an existing code 4. Advanced
More informationThreaded Programming. Lecture 9: Alternatives to OpenMP
Threaded Programming Lecture 9: Alternatives to OpenMP What s wrong with OpenMP? OpenMP is designed for programs where you want a fixed number of threads, and you always want the threads to be consuming
More informationHPC with Multicore and GPUs
HPC with Multicore and GPUs Stan Tomov Electrical Engineering and Computer Science Department University of Tennessee, Knoxville COSC 594 Lecture Notes March 22, 2017 1/20 Outline Introduction - Hardware
More informationStarPU: a runtime system for multigpu multicore machines
StarPU: a runtime system for multigpu multicore machines Raymond Namyst RUNTIME group, INRIA Bordeaux Journées du Groupe Calcul Lyon, November 2010 The RUNTIME Team High Performance Runtime Systems for
More informationCUDA Toolkit 5.0 Performance Report. January 2013
CUDA Toolkit 5.0 Performance Report January 2013 CUDA Math Libraries High performance math routines for your applications: cufft Fast Fourier Transforms Library cublas Complete BLAS Library cusparse Sparse
More informationExperiences Using Tegra K1 and X1 for Highly Energy Efficient Computing
Experiences Using Tegra K1 and X1 for Highly Energy Efficient Computing Gaurav Mitra Andrew Haigh Luke Angove Anish Varghese Eric McCreath Alistair P. Rendell Research School of Computer Science Australian
More informationOvercoming the Barriers to Sustained Petaflop Performance. William D. Gropp Mathematics and Computer Science
Overcoming the Barriers to Sustained Petaflop Performance William D. Gropp Mathematics and Computer Science www.mcs.anl.gov/~gropp But First Are we too CPU-centric? What about I/O? What do applications
More information