Fabulous book, it connects what you learned in school to real life situations. Even after years in the industry, unless you are on the forefront of performance optimization many of these knowledge are gonna help you. In theory I have learned all these in school but school knowledge here has not connect with industry experiences and this book did it.
Download the free Kindle app and start reading Kindle books instantly on your smartphone, tablet, or computer - no Kindle device required.
Read instantly on your browser with Kindle for Web.
Using your mobile phone camera - scan the code below and download the Kindle app.
Customers who read this book also read
Page 1 of 1 Start overPage 1 of 1
From the brand
-
-
Packt is a leading publisher of technical learning content with the ability to publish books on emerging tech faster than any other.
Our mission is to increase the shared value of deep tech knowledge by helping tech pros put software to work.
We help the most interesting minds and ground-breaking creators on the planet distill and share the working knowledge of their peers.
From the Publisher
How has your background helped you to write this book?
I started programming as a college student when even the simplest programs had to be written with performance in mind. While times have changed, I kept finding myself working in the areas where the hardware capabilities were never enough: first, in computational physics and then in electronic design automation. By choice or circumstance, I have always been programming close to hardware - the skill that is in demand once more in recent years in areas like low latency and high-performance concurrency. As for compilers, their ability to improve performance is miles ahead of what it once was, and so is the potential magnitude of the failure when the expected performance boost does not materialize due to a quirk of the language or a peculiarity of the hardware interfering with the programmer's intent. All in all, I found the old habits, if not the old knowledge, relevant again in the modern age of high-performance computing.
What part or parts of the book are your favorite, and why?
Several parts, and the experience of writing them, stand out in my memory. First of all, it's the chapter on designing for performance. So often, performance is considered a feature to be added later. The mantra "do not optimize prematurely" is as widely known as it is misunderstood; its intended meaning is much more nuanced than most programmers realize. The idea that performance targets are as much a part of the original design specification as the feature list often comes as a surprise; if I can teach one thing to the readers of my book, I would want it to be this idea.
Other favorite moments are the ones where I found and presented some minor or innocuous change that has an unexpectedly large effect on performance. Certain compiler optimizations, seemingly trivial changes to memory access, and clever concurrent algorithms all have that element of surprise and even disbelief: "It does WHAT?"
In what ways have you been active in the community?
Mostly in various educational activities, like conference presentations and CppCon Academy classes. Sharing my knowledge and having the knowledge and experience of others shared with me are where I find my role in the C++ community most useful.
What makes this book better than other similar titles?
Obviously, the content, the presentation, and the style! Seriously though, there are several strengths of the book in my mind. First is the emphasis on not just imparting knowledge and information, but explaining how we came to know what we know, and how such knowledge can be re-discovered when the next generations of hardware, languages, and compilers will make the guidelines of today obsolete. In this way, it is a book about learning to learn. Second is the scope, the breadth of it: it is rare to find, in one place, the presentation of the performance effects that originate from the hardware, the compilers, the language features, and the design decisions, and their mutual interactions.
About the author
Excellent read for improving program runtimes at the CPU level
The book “The art of writing efficient programs by Fedor G. Pikusis for experienced developers and programmers who want to learn different methods to improve the performance of their code. The developers who are involved in computer modeling, algorithmic trading and genomics can gain a lot from this book.The first section of the book describes the progress that computing power has made over the years from taking 10 hrs to solve problems in 2008 to less than 1 hour in 2018 and even faster today in 2021. So all this optimization is because of improvement in technologies in software and optimized algorithms.In chapter 1 the author introduces performance and efficiency. Efficient programs make good use of the computational hardware. High performance requires choosing the right CPU, writing the code to execute the program, avoiding unnecessary computations. In chapter 2 the author discusses performance measurements. The first rule is never guess about performance. This chapter describes the tool google-pprof that times each system call that the compiler takes to process. The perf profiler, and C++ chrono timers are described in this section. In chapter 3 the author describes how to make use of CPU resources effectively. In chapter 4 the author discusses memory optimization and concurrency. The spectre error is discussed in memory optimization. The use of mutexes to control the critical section of the code is important to determine the correct order in which the criticl section of the code is serviced. In chapter 5 threads, memory and concurrency a detailed discussion od multi threading fundamentals is done. In chapter 6 concurrency and performance the discussion of parallel programs and performing lock free optimizations are discussed. In chapter 7 data structures for concurrency are discussed.Chapter 8 discusses concurrency in C++Chapter 9 discusses high performance optimizations in C++. Chapter 10 discusses compiler optimizations for C++. Avoiding unnecessary memory allocation and memory fragmentation in programs. Allocate memory in chunks of 64KB. Chapter 11 discusses undefined behaviour and performance. undefined behaviour happens when the program behaves in a manner different from that which is specified. Chapter 12 discuess design for performance.Overall this book is highly recommended for CPU enthusiasts and programmers in the area of high performance computing. When speed is all that matters for solving critical problems this is an excellent resource to use.
Top reviews from the United States
There was a problem filtering reviews right now. Please try again later.
Reviewed in the United States on August 26, 2024
Reviewed in the United States on October 23, 2022
Reviewed in the United States on May 12, 2024
The book's layout is poorly designed. For example, there's an unnecessary gray background in the source code sections, which could have been improved by highlighting only the important parts with a gray background and using bold for keywords. While the content of the book is quite good and interesting to read, the layout is frustrating.
Images in this review
Reviewed in the United States on November 17, 2021
Reviewed in the United States on February 11, 2023
Top reviews from other countries
Ory Chowaw-Liebman
Highly inspiring book for C++ developers.
Reviewed in Germany on February 14, 2022Klaus Iglberger
Good Introduction to the essential performance aspects for aspiring C++ developers
Reviewed in Germany on December 14, 2021Augusto H.
The graphics are unreadable, and that really hurts understanding
Reviewed in Germany on November 19, 2022
Augusto H.
Reviewed in Germany on November 19, 2022
Images in this review
