Michael Wolfe's Programming Guide
The PGI Accelerator Programming Model on NVIDIA GPUs—Part 1
GPUs have a very high compute capacity, and recent designs have made them more programmable and useful for tasks other than just graphics. This article is the first in a series on PGI's accelerator-enabled compilers. Part 1 takes an in-depth look at PGI's Accelerator Programming Model.
Also in this issue…
Tool Tips: Accessing Compiler Performance Advice
Traditional methods of performance tuning can only get you so far. Using PGI compiler feedback with the PGPROF profiler can ease the task of improving application performance. Part 1 in a series.
Cross Platform: Developing Local MPI Programs
Systems with PGI Workstation installed can build, run, debug and profile MPI applications with up to eight processes running locally. This article describes how to set up and go about developing MPI applications locally placing particular emphasis on differences between Linux, Mac OS X and Windows.
An Accelerator Programming Model for Multicore
This article explores using a version of the PGI Accelerator Programming Model on a multicore target. The authors attempt to answer two questions: whether the model can target a multicore processor, and whether doing so is a good idea.
Upcoming Events
ISC'09, Hamburg, 23-26 June, booth A2.3. PGI will also be hosting "A High Level GPU/Accelerator Programming Model" BoF at 12:20 PM Friday in Hall C2.2.
Michael Wolfe will be presenting a one hour webinar titled "PGI x64+GPU Fortran & C99 Compilers" on the PGI Accelerator programming model on June 16. Hosted by IEEE Spectrum. Register to attend.
New Resources
PGI Accelerator Compiler Overview video
PGI Accelerator Compiler Examples video
PGI Fortran & C Accelerator Programming Model whitepaper
OpenMP 3.0 and Task Parallelization video
Migrating HPC Applications from UNIX to Windows whitepaper
Tools for the Classic HPC Developer whitepaper
Next Issue Highlights
PGI Accelerator Programming Model—Part 2: Tuning
New Sample-based Profiling Feature in PGPROF 9.0
Using Interprocedural Analysis
Debugging Remote MSMPI Programs with PVF
Release 9.0/Accelerator Road Map
You can borrow a seat of your network floating license for use when your laptop or notebook is disconnected from the license server. More information is available on the lmborrow description and instructions page at pgroup.com.
- The Portland Group, Inc.
- Suite 320
- Two Centerpointe Drive
- Lake Oswego, OR 97035