Downloaded 27 times
![8
OPENACC DIRECTIVES
Manage
Data
Movement
Initiate
Parallel
Execution
Optimize
Loop
Mappings
#pragma acc data copyin(a,b) copyout(c)
{
...
#pragma acc parallel
{
#pragma acc loop gang vector
for (i = 0; i < n; ++i) {
c[i] = a[i] + b[i];
...
}
}
...
}
• CPU, GPU, Manycore
• Performance portable
• Interoperable
• Single source base
• Incremental](https://image.slidesharecdn.com/pgiwwfoupdate18-180313225819/75/PGI-Compilers-Tools-Update-March-2018-8-2048.jpg)
![22
#pragma acc data copyin(a,b) copyout(c)
{
...
#pragma acc parallel
{
#pragma acc loop gang vector
for (i = 0; i < n; ++i) {
c[i] = a[i] + b[i];
...
}
}
...
}
PGI OpenACC and CUDA Unified Memory
Compiling with the –ta=tesla:managed option
C malloc, C++ new, Fortran allocate all mapped to CUDA Unified Memory
GPU Developer View With
CUDA Unified Memory
Unified Memory](https://image.slidesharecdn.com/pgiwwfoupdate18-180313225819/75/PGI-Compilers-Tools-Update-March-2018-22-2048.jpg)
![23
PGI OpenACC and CUDA Unified Memory
Compiling with the –ta=tesla:managed option
C malloc, C++ new, Fortran allocate all mapped to CUDA Unified Memory
...
#pragma acc parallel
{
#pragma acc loop gang vector
for (i = 0; i < n; ++i) {
c[i] = a[i] + b[i];
...
}
}
...
GPU Developer View With
CUDA Unified Memory
Unified Memory](https://image.slidesharecdn.com/pgiwwfoupdate18-180313225819/75/PGI-Compilers-Tools-Update-March-2018-23-2048.jpg)
Uploaded byNVIDIA
PGI Compilers & Tools Update- March 2018
This document provides an update on PGI compilers and tools for heterogeneous supercomputing. It discusses PGI's support for OpenACC directives to accelerate applications on multicore CPUs and NVIDIA GPUs from a single source. It highlights new compiler features including support for Intel Skylake, AMD EPYC and IBM POWER9 CPUs as well as NVIDIA Volta GPUs. Benchmark results show strong performance of OpenACC applications on these platforms. The document also discusses the growing adoption of OpenACC in HPC applications and resources available to support OpenACC development.
More Related Content
Similar to PGI Compilers & Tools Update- March 2018
![[01][gpu 컴퓨팅을 위한 언어, 도구 및 api] miller languages tools](https://cdn.slidesharecdn.com/ss_thumbnails/01gpuapimillerlanguagestools-110106231409-phpapp01-thumbnail.jpg?width=640&height=640&fit=bounds)
PGI Compilers & Tools Update- March 2018
- 1. PGI Compilers forHeterogeneous Supercomputing, March 2018 PGI COMPILERS & TOOLS UPDATE
- 2. 2 PGI - THENVIDIA HPC SDK Fortran, C & C++ Compilers Optimizing, SIMD Vectorizing, OpenMP Accelerated Computing Features OpenACC Directives CUDA Fortran Multi-Platform Solution Multicore x86-64 and OpenPOWER CPUs, NVIDIA Tesla GPUs Supported on Linux, macOS, Windows MPI/OpenMP/OpenACC Tools Debugger Performance Profiler Interoperable with DDT, TotalView
- 3. 3 OPENACC FOR EVERYONE PGICommunity Edition Now Available PROGRAMMING MODELS OpenACC, CUDA Fortran, OpenMP, C/C++/Fortran Compilers and Tools PLATFORMS X86, OpenPOWER, NVIDIA GPU UPDATES 1-2 times a year 6-9 times a year 6-9 times a year SUPPORT User Forums PGI Support PGI Premier Services LICENSE Annual Perpetual Volume/Site FREE
- 4. 4 Latest CPUs Support IntelSkylake AMD Zen IBM POWER9 Full OpenACC 2.6 OpenMP 4.5 for multicore CPUs AVX-512 code generation Integrated CUDA 9.1 toolkit/libraries New fastmath intrinsics library Partial C++17 support Optional LLVM-based x86 code generator pgicompilers.com/whats-new
- 5. 5 Performance measured February,2018. Skylake: Two 20 core Intel Xeon Gold 6148 CPUs @ 2.4GHz w/ 376GB memory, hyperthreading enabled. EPYC: Two 24 core AMD EPYC 7451 CPUs @ 2.3GHz w/ 256GB memory. Broadwell: Two 20 core Intel Xeon E5-2698 v4 CPUs @ 3.6GHz w/ 256GB memory, hyperthreading enabled. Volta: NVIDIA DGX1 system with two 20 core Intel Xeon E5-2698 v4 CPUs @ 2.20GHz, 256GB memory, one NVIDIA Tesla V100-SXM2-16GB GPU @ 1.53GHz. SPEC® is a registered trademark of the Standard Performance Evaluation Corporation (www.spec.org). SPEC ACCEL 1.2 BENCHMARKS 0 50 100 150 200 2-socket Skylake 2-socket EPYC 2-socket BroadwellGEOMEANSeconds Intel 2018 PGI 18.1 OpenMP 4.5 40 cores / 80 threads 48 cores / 48 threads 40 cores / 80 threads 0 50 100 150 200 GEOMEANSeconds PGI 18.1 OpenACC 2-socket Broadwell 1x Volta V100 4.4x Speed-up
- 6. 6 0 50 100 150 200 2-socket Skylake 2-socketEPYC 2-socket Broadwell GEOMEANSeconds Intel 2018 PGI 18.1 SPEC CPU 2017 FP SPEED BENCHMARKS 40 cores / 80 threads 48 cores / 48 threads 40 cores / 80 threads Performance measured February, 2018. Skylake: Two 20 core Intel Xeon Gold 6148 CPUs @ 2.4GHz w/ 376GB memory, hyperthreading enabled. EPYC: Two 24 core AMD EPYC 7451 CPUs @ 2.3GHz w/ 256GB memory. Broadwell: Two 20 core Intel Xeon E5-2698 v4 CPUs @ 3.6GHz w/ 256GB memory, hyperthreading enabled. SPEC® is a registered trademark of the Standard Performance Evaluation Corporation (www.spec.org).
- 7.
- 8. 8 OPENACC DIRECTIVES Manage Data Movement Initiate Parallel Execution Optimize Loop Mappings #pragma accdata copyin(a,b) copyout(c) { ... #pragma acc parallel { #pragma acc loop gang vector for (i = 0; i < n; ++i) { c[i] = a[i] + b[i]; ... } } ... } • CPU, GPU, Manycore • Performance portable • Interoperable • Single source base • Incremental
- 9. 9 CPU OPENACC IS FORMULTICORE CPUS & GPUS % pgfortran -ta=multicore –fast –Minfo=acc -c update_tile_halo_kernel.f90 . . . 100, Loop is parallelizable Generating Multicore code 100, !$acc loop gang 102, Loop is parallelizable GPU % pgfortran -ta=tesla –fast -Minfo=acc –c update_tile_halo_kernel.f90 . . . 100, Loop is parallelizable 102, Loop is parallelizable Accelerator kernel generated Generating Tesla code 100, !$acc loop gang, vector(4) ! blockidx%y threadidx%y 102, !$acc loop gang, vector(32) ! blockidx%x threadidx%x 98 !$ACC KERNELS 99 !$ACC LOOP INDEPENDENT 100 DO k=y_min-depth,y_max+depth 101 !$ACC LOOP INDEPENDENT 102 DO j=1,depth 103 density0(x_min-j,k)=left_density0(left_xmax+1-j,k) 104 ENDDO 105 ENDDO 106 !$ACC END KERNELS
- 10. 10 0 20 40 60 80 100 120 140 160 Multicore Haswell MulticoreBroadwell Multicore Skylake Kepler Pascal PGI 18.1 OpenACC Intel 2018 OpenMP 7.6x 7.9x 10x 10x 11x 40x 14.8x 15x Volta V100 CLOVERLEAF SpeedupvsSingleHaswellCore Systems: Haswell: 2x16 core Haswell server, four K80s, CentOS 7.2 (perf-hsw10), Broadwell: 2x20 core Broadwell server, eight P100s (dgx1-prd-01), Broadwell server, eight V100s (dgx07), Skylake 2x20 core Xeon Gold server (sky-4). Compilers: Intel 2018.0.128, PGI 18.1 Benchmark: CloverLeaf v1.3 downloaded from http://uk-mac.github.io/CloverLeaf the week of November 7 2016; CloverlLeaf_Serial; CloverLeaf_ref (MPI+OpenMP); CloverLeaf_OpenACC (MPI+OpenACC) Data compiled by PGI February 2018. AWE Hydrodynamics mini-App, bm32 data set http://uk-mac.github.io/CloverLeaf 109x 67x 142x 1x 2x 4x
- 11. 11 OPENACC UPTAKE INHPC Applications 3 of Top 5 HPC Apps ANSYS Fluent & Gaussian released; VASP in development 5 ORNL CAAR Codes GTC, XGC, ACME, FLASH, LSDalton 109 Apps Total being tracked Hackathons Training Community 5 Events in 2017 All hackathons are initiated by users 6–9 Events in 2018 94 Codes GPU accelerated to date Expertise 94 mentors registered User Group SC17 participation up 27% vs SC16 Slack Channel 2x growth in last 6 months Downloads PGI Community Edition quarterly downloads up 136% in 2017 OpenACC and DLI 10 new modules; instructor certification 18+ 2018 workshops ECMWF, KAUST, PSC, CESGA Online Courses 5K+ attended over last 3 years Online Labs 4.3K+ taken over last 3 years
- 12. 12 GAUSSIAN 16 Using OpenACCallowed us to continue development of our fundamental algorithms and software capabilities simultaneously with the GPU-related work. In the end, we could use the same code base for SMP, cluster/ network and GPU parallelism. PGI's compilers were essential to the success of our efforts. Mike Frisch, Ph.D. President and CEO Gaussian, Inc. Gaussian, Inc. 340QuinnipiacSt. Bldg. 40 Wallingford, CT 06492USA [email protected] Gaussian isa registered trademark of Gaussian, Inc. All other trademarksand thepropertiesof their respectiveholders. Specif cationssubject tochangewitho Copyright © 2017, Gaussian, Inc. All rightsreserved. Roberto Gomperts NVIDIA Michael Frisch Gaussian Brent Leback NVIDIA/PGI Gio Project Contributors %GPUCPU=0 - 7 =0 - 7 UseGPUs0-7with CPUs0-7astheir controllers. Detailed information isavailableon our website.
- 13. 13 ANSYS FLUENT We’ve effectivelyused OpenACC for heterogeneous computing in ANSYS Fluent with impressive performance. We’re now applying this work to more of our models and new platforms. Sunil Sathe Lead Software Developer ANSYS Fluent
- 14. 14 VASP For VASP, OpenACCis the way forward for GPU acceleration. Performance is similar and in some cases better than CUDA C, and OpenACC dramatically decreases GPU development and maintenance efforts. We’re excited to collaborate with NVIDIA and PGI as an early adopter of CUDA Unified Memory. Prof. Georg Kresse Computational Materials Physics University of Vienna
- 15. 15 MPAS-A Our team hasbeen evaluating OpenACC as a pathway to performance portability for the Model for Prediction (MPAS) atmospheric model. Using this approach on the MPAS dynamical core, we have achieved performance on a single P100 GPU equivalent to 2.7 dual socketed Intel Xeon nodes on our new Cheyenne supercomputer. Richard Loft Director, Technology Development NCAR Image courtesy: NCAR
- 16. 16 David Gutzwiller Lead SoftwareDeveloper NUMECA NUMECA FINE/Open Porting our unstructured C++ CFD solver FINE/Open to GPUs using OpenACC would have been impossible two or three years ago, but OpenACC has developed enough that we’re now getting some really good results.
- 17. 17 OpenACC made itpractical to develop for GPU-based hardware while retaining a single source for almost all the COSMO physics code. Dr. Oliver Fuhrer Senior Scientist Meteoswiss COSMO
- 18. 18 GAMERA FOR GPU WithOpenACC and a compute node based on NVIDIA's Tesla P100 GPU, we achieved more than a 14X speed up over a K Computer node running our earthquake disaster simulation code Takuma Yamaguchi, Kohei Fujita, Tsuyoshi Ichimura, Muneo Hori, Lalith Wijerathne The University of Tokyo Map courtesy University of Tokyo
- 19. 19 QUANTUM ESPRESSO CUDA Fortrangives us the full performance potential of the CUDA programming model and NVIDIA GPUs. !$CUF KERNELS directives give us productivity and source code maintainability. It’s the best of both worlds. Filippo Spiga Head of Research Software Engineering University of Cambridge
- 20. 20 OPENACC AND CUDAUNIFIED MEMORY
- 21. 21 Programming GPU-Accelerated Systems CUDAUnified Memory for Dynamically Allocated Data GPU Developer View With CUDA Unified Memory Unified Memory GPU Developer View System Memory GPU Memory PCIe
- 22. 22 #pragma acc datacopyin(a,b) copyout(c) { ... #pragma acc parallel { #pragma acc loop gang vector for (i = 0; i < n; ++i) { c[i] = a[i] + b[i]; ... } } ... } PGI OpenACC and CUDA Unified Memory Compiling with the –ta=tesla:managed option C malloc, C++ new, Fortran allocate all mapped to CUDA Unified Memory GPU Developer View With CUDA Unified Memory Unified Memory
- 23. 23 PGI OpenACC andCUDA Unified Memory Compiling with the –ta=tesla:managed option C malloc, C++ new, Fortran allocate all mapped to CUDA Unified Memory ... #pragma acc parallel { #pragma acc loop gang vector for (i = 0; i < n; ++i) { c[i] = a[i] + b[i]; ... } } ... GPU Developer View With CUDA Unified Memory Unified Memory
- 24. 24NVIDIA CONFIDENTIAL. DONOT DISTRIBUTE. GTC: An OpenACC Production Application The gyrokinetic toroidal code (GTC) is a massively parallel, particle-in-cell production code for turbulence simulation in support of the burning plasma experiment ITER, the crucial next step in the quest for fusion energy. Being ported for runs on the ORNL Summit supercomputer http://phoenix.ps.uci.edu/gtc_group
- 25. 25 GTC Performance usingOpenACC P8 : IBM POWER8NVL, 2 sockets, 20 cores, NVLINK UM : No Data Directives in sources, compiled with –ta=tesla:managed 2x 4x 6x 8x 10x 12x OpenPOWER | NVLink | Unified Memory | P100 | V100 14x 16x Data Directives Data Directives Data Directives 6.1X 5.9X 12.1X 12X 16.5X 20-core P8 P8+2xP100 UM P8+2xP100 P8+4xP100 UM P8+4xP100 x64+4xV100
- 26. 26 Resources https://www.openacc.org/resources Success Stories https://www.openacc.org/success-stories Events https://www.openacc.org/events OPENACC RESOURCES Guides● Talks ● Tutorials ● Videos ● Books ● Spec ● Code Samples ● Teaching Materials ● Events ● Success Stories ● Courses ● Slack ● Stack Overflow Compilers and Tools https://www.openacc.org/tools www.openacc.org/community#slack PGI User Forums pgicompilers.com/userforum stackoverflow.com/questions/tagged /openacc' Questions Support Options