Multiple Onsite Events at SC16 in Salt Lake City to Showcase Recent Advances BEND, OREGON– NOVEMBER 7, 2016 – The OpenACC standards group today announced several major milestones including the addition of new member, the National Supercomputing Center in Wuxi, the adoption of OpenACC by several major HPC applications, the addition of support for new target platforms and … [Read more...]
OpenACC Highlights at GTC 2016
GTC 2016, the upcoming GPU Technology Conference (GTC), which is being held in San Jose, CA April 4-7, 2016 contains several OpenACC talks of note. If you are attending, check out the following OpenACC related activities. If not, NVIDIA will put the videos online within a short period after the conference. S6524 - Enabling the Electronic Structure Program Gaussian on GPGPUs … [Read more...]
OpenACC Announcements Including PGI Compiler Free to Academics and Students!
OpenACC Standards Group Expands With New Members and Education Organization Extends Developer Momentum Approaching SC 15 While Planning More Hackathons In 2016 BEND, OR-NOVEMBER 9, 2015 – The OpenACC standards group today announced the addition of two new members, recent success stories from OpenACC applications and the schedule for Hackathons and Training Workshops in the … [Read more...]
Excellent x86 Performance! – PGI Accelerator Compilers Add OpenACC Support for x86 Multicore CPUs
TechEnablement - OpenACC x86 multi-core performance matches that of the Intel compiler using OpenMP when using the latest 15.10 version of the Portland Group compiler. Note: two MPI instances per GPU were required to fully utilize the NVIDIA K80 GPU as the problem sizes were too small to fully utilize the GPU with a single instance. SANTA CLARA, Calif.—Oct. 29, … [Read more...]
The First of NVIDIA’s Free OpenACC Courses is Online!
The first of NVIDIA's free online OpenACC course (Download / Stream, Slides,Q&A), lab (in C/C++ or Fortran), and online "office hours" (Download / Stream, Slides) is available to everyone! OpenACC is a programming approach designed to minimize developer effort while delivering performance portability on x86 CPUs and GPUs. The 8-week online OpenACC course described on … [Read more...]
Free Online OpenACC Course Starting Oct. 1 2015
NVIDIA is providing a free, interactive, online OpenACC course starting on October 1, 2015. The course is made up of four instructor-led classes that include interactive lectures, hands-on exercises, and live office hours with the instructors. Register here! Classes start at 9am PT on Thursdays and "Office Hours" are 9AM PT on Tuesdays. All sessions will be recorded for … [Read more...]
PathScale Supports OpenMP4 and OpenACC On NVIDIA, AMD, ARMv8, Plus Some CUDA
PathScale is capitalizing on over six years of development with the release of ENZO 2015, a compiler suite that allows C/C++/Fortran parallel programs to be built for a variety of hardware platforms from a single OpenACC or OpenMP4 source base. The PathScale compilers can generate executables for AMD GPUs, and NVIDIA GPUs as well as ARMv8 and x86 instruction sets. As a result, … [Read more...]
PGI Compiled OpenACC ILP Loop Beats CUDA-7 by 200 GF/s on Deep-learning PCA Example
The PGI OpenACC compiler beat the performance of a CUDA 7.0 NVIDIA nvcc compiled deep-learning based PCA (Principal Components Analysis) example by 200 GF/s on a K40c using an ILP (Instruction Level Parallelism) loop structure taught in the TechEnablement classes and forthcoming Farber OpenACC book. PCA is an important data analysis tool utilized by data scientists. Sign up for … [Read more...]
GCC5 Release Candidate in Early April With OpenACC, Cilk, and OpenMP4 Offload!
Richard Biener posted a message to gcc.gnu.org to expect a GCC 5 release candidate the first week of April. GCC5 supports OpenACC, Cilk Plus, and OpenMP 4.0 offload support. "We've come a long way towards the release criteria of zero P1 bugs. There are still a few remaining P1s though and we are targeting for a GCC 5 release candidate in the first week of April (given those … [Read more...]
Concise Comparision Adds OpenMP Versus OpenACC To CUDA Versus OpenCL Debates
The debate over OpenMP versus OpenACC for manycore and heterogeneous computing is starting to heat up. Michael Wong (CEO of OpenMP Corp), Barbara Chapman (Univ. Houston), and Yonghong Yan (Assistant Prof. Univ. Oakland and OpenMP ARB representative) have written a nice, quick read, comparative article on HPCWire: "A Comparison of Heterogeneous and Manycore … [Read more...]