High-Performance GPU Computing provides a practical exploration of designing scalable parallel systems for accelerated and distributed workloads. The book explains how modern GPU architectures, parallel programming models, memory hierarchies, workload scheduling, and high-speed interconnects can be used to improve computational performance and system efficiency.
Readers will gain insights into GPU-accelerated application development, multi-GPU processing, distributed computing frameworks, performance profiling, workload optimization, fault tolerance, and resource management. The book also examines the challenges of scaling compute-intensive applications across clusters, cloud platforms, and heterogeneous infrastructure.
Designed for software engineers, system architects, researchers, data scientists, and technology professionals, this book connects fundamental parallel computing concepts with practical strategies for building reliable, efficient, and high-performance computing environments. It offers a structured foundation for developing accelerated systems capable of supporting scientific computing, artificial intelligence, data analytics, simulation, visualization, and other demanding workloads.
