PGI Compilers & Tools are used by scientists and engineers developing applications for high-performance computing (HPC). PGI products deliver world-class…
Overview
The article announces the release of the PGI Community Edition version 17.10, highlighting its features and improvements for high-performance computing applications. It emphasizes support for Tesla V100 GPUs, OpenACC for CUDA Unified Memory, and various performance optimizations.
What You'll Learn
1
How to utilize OpenACC for CUDA Unified Memory in your applications
2
Why to adopt PGI Compilers for multicore CPU performance
3
When to implement C++14 lambdas with capture in OpenACC regions
Key Questions Answered
What new features are included in PGI 17.10?
PGI 17.10 introduces several new features including support for Tesla V100 GPUs, OpenACC for CUDA Unified Memory, OpenMP 4.5 for multicore CPUs, C++14 lambdas with capture in OpenACC regions, and performance optimizations for C++. These enhancements aim to improve performance and usability for high-performance computing applications.
How does PGI Compilers support GPU computing?
PGI Compilers provide an easy on-ramp to GPU computing through OpenACC directives, allowing developers to leverage GPU capabilities without extensive knowledge of GPU programming. This facilitates better performance portability across major HPC platforms.
Technologies & Tools
Software
Pgi Compilers & Tools
Used for developing applications in high-performance computing environments.
Framework
Openacc
Provides directives for GPU computing.
Framework
Openmp
Supports multicore CPU programming.
Key Actionable Insights
1Adopting PGI Compilers can significantly enhance application performance on multicore CPUs and GPUs.By utilizing the latest features in PGI 17.10, developers can optimize their applications for high-performance computing environments, which is crucial for resource-intensive tasks.
2Utilizing OpenACC for CUDA Unified Memory can simplify memory management in GPU applications.This feature allows developers to focus more on algorithm development rather than low-level memory handling, streamlining the development process.
3Incorporating C++14 lambdas with capture in OpenACC regions can lead to cleaner and more efficient code.This modern C++ feature enhances code readability and maintainability, which is particularly beneficial in complex HPC applications.
Common Pitfalls
1
Failing to leverage the latest features of PGI Compilers can lead to suboptimal application performance.
Many developers may stick to older versions or methods, missing out on significant performance improvements available in newer releases.