menu
{ "item_title" : "CUDA 13.4 Performance Engineering", "item_author" : [" Stevenson Ortega "], "item_description" : "Ready to take your GPU programming skills beyond simply writing kernels that work? What if you could understand why a kernel performs poorly, identify where the bottleneck comes from, and systematically improve its speed and efficiency? CUDA 13.4 Performance Engineering is a practical guide for developers, engineers, students, and technical professionals who want to develop a deeper understanding of GPU performance and modern parallel computing. Instead of focusing only on syntax or basic programming concepts, this book explores the reasoning behind high-performance GPU applications and shows you how to approach optimization as an engineering discipline. You will explore GPU kernel optimization, thread organization, execution models, memory hierarchies, memory access patterns, synchronization, occupancy, parallel execution, and performance trade-offs. How much performance are you leaving on the table because of inefficient memory access? Could a different thread or block organization improve your workload? What happens when computation becomes limited by memory bandwidth rather than arithmetic throughput? The book helps you answer questions like these while developing a structured approach to performance analysis. You will also learn how to use profiling and benchmarking techniques to locate bottlenecks, interpret performance measurements, evaluate optimization strategies, and avoid changes that appear beneficial but provide little real-world improvement. Practical discussions connect low-level GPU behavior with demanding workloads such as numerical computing, data processing, scientific applications, machine learning, and AI acceleration. Whether you are beginning to explore GPU performance engineering or looking to strengthen an existing optimization workflow, this book provides a focused foundation for understanding how parallel hardware, software design, memory behavior, and computational workloads interact. By the end, you will have a clearer framework for designing, analyzing, profiling, and improving GPU-accelerated applications with greater confidence and precision.", "item_img_path" : "https://covers4.booksamillion.com/covers/bam/9/79/816/560/9798165605727_b.jpg", "price_data" : { "retail_price" : "37.99", "online_price" : "37.99", "our_price" : "37.99", "club_price" : "37.99", "savings_pct" : "0", "savings_amt" : "0.00", "club_savings_pct" : "0", "club_savings_amt" : "0.00", "discount_pct" : "10", "store_price" : "" } }
CUDA 13.4 Performance Engineering|Stevenson Ortega

CUDA 13.4 Performance Engineering : Master GPU Kernel Optimization, Memory Hierarchies, Parallel Execution, Profiling, and AI Acceleration

local_shippingShip to Me
In Stock.
FREE Shipping for Club Members help

Overview

Ready to take your GPU programming skills beyond simply writing kernels that work? What if you could understand why a kernel performs poorly, identify where the bottleneck comes from, and systematically improve its speed and efficiency? CUDA 13.4 Performance Engineering is a practical guide for developers, engineers, students, and technical professionals who want to develop a deeper understanding of GPU performance and modern parallel computing. Instead of focusing only on syntax or basic programming concepts, this book explores the reasoning behind high-performance GPU applications and shows you how to approach optimization as an engineering discipline. You will explore GPU kernel optimization, thread organization, execution models, memory hierarchies, memory access patterns, synchronization, occupancy, parallel execution, and performance trade-offs. How much performance are you leaving on the table because of inefficient memory access? Could a different thread or block organization improve your workload? What happens when computation becomes limited by memory bandwidth rather than arithmetic throughput? The book helps you answer questions like these while developing a structured approach to performance analysis. You will also learn how to use profiling and benchmarking techniques to locate bottlenecks, interpret performance measurements, evaluate optimization strategies, and avoid changes that appear beneficial but provide little real-world improvement. Practical discussions connect low-level GPU behavior with demanding workloads such as numerical computing, data processing, scientific applications, machine learning, and AI acceleration. Whether you are beginning to explore GPU performance engineering or looking to strengthen an existing optimization workflow, this book provides a focused foundation for understanding how parallel hardware, software design, memory behavior, and computational workloads interact. By the end, you will have a clearer framework for designing, analyzing, profiling, and improving GPU-accelerated applications with greater confidence and precision.

This item is Non-Returnable

Details

  • ISBN-13: 9798165605727
  • ISBN-10: 9798165605727
  • Publisher: Independently Published
  • Publish Date: October 2026
  • Dimensions: 11 x 8.5 x 0.79 inches
  • Shipping Weight: 1.96 pounds
  • Page Count: 384

Related Categories

You May Also Like...

    1

BAM Customer Reviews