Parallel and High Performance Computing

Viac o knihe

Parallel and High Performance Computing provides essential techniques to enhance your code’s efficiency. Complex tasks, such as training deep learning models or executing large-scale simulations, can be time-consuming. By implementing efficient parallel programming, you can significantly reduce computing time, sometimes saving hours or even days. This guide teaches you to achieve faster run-times, improved scalability, and enhanced energy efficiency by leveraging parallel techniques for multicore processors and GPU hardware. You will learn to write fast, powerful, and energy-efficient programs capable of handling vast data volumes. The book covers evaluating hardware architectures and utilizing industry-standard tools like OpenMP and MPI. You'll master data structures and algorithms tailored for high-performance computing and discover methods to conserve energy on handheld devices. It even includes a practical example of running a massive tsunami simulation across multiple GPUs. Designed for experienced programmers familiar with high-performance computing languages such as C, C++, or Fortran, the content includes planning parallel projects, understanding CPU and GPU architecture differences, addressing performance issues, and managing applications with batch scheduling. The authors, Robert Robey and Yuliana Zamora, bring extensive expertise in parallel computing and programming modern hardware.