Bicho de sete cabeças. It's not a seven-headed beast. It's not rocket science.Cassio Neri Fast conversion from floating-point numbers CppCon 2024 - Aurora (Independent Researcher)MeMe O Neri's equationMe

Andreas Fertig Write unique code! Fast and small C++ When efficiency matters Presentation Material CppCon, Aurora CO, 2024-09-17© 2024 Andreas Fertig AndreasFertig.com All rights reserved All programs Fertig v1.0 Fast and small C++ 2 Andreas Fertig v1.0 Fast and small C++ 3 Fast and small C++ © 2024 Andreas Fertig https://AndreasFertig.com post@AndreasFertig.com 4Andreas Fertig v1.0 Fast and small std::unique_ptr{fopen("SomeFile.txt", "r"), &fclose}; Andreas Fertig v1.0 Fast and small C++ 5 Fast and small C++ © 2024 Andreas Fertig https://AndreasFertig.com post@AndreasFertig.com

C++ still does not have reflection (introspection): we have to “pick our poison”: • Doing the conversion manually (good luck, you’ll need it) • Automating: • Preprocessing (my poison of choice, twice

Making Games Start Fast A Story About ConcurrencyOnce upon a time, in the early 2000s...We had a small clock speed cap issue...But the solution was easy! Core 🔥🔥🔥 Core 1 Core 2 Core 3 Core 4The Data has to be loaded, but does it have to loaded now? ◉ Our goal is to display the main menu as fast as possible ◉ We can continue loading in background 65Side by Side Comparison 66Loading audio

RapidJSON Boost.JSON simdjson 0 10 20 30 40 Library Time in ms fully validating pull parser fast, O(1) memory usage implicitly define schema github.com/think-cell/think-cell-library Jonathan Müller "Jonathan Müller", "speaker": true, "talks": [ { "title": "Express your expectations", "subtitle": "A fast, compliant JSON pull parser..." } ] } Jonathan Müller — @foonathan Express your expectations CppCon log(object.name) console.log(object.talks[0].subtitle) console.log(object.foo) Jonathan Müller A fast, compliant JSON pull parser for writing robust applications undefined Jonathan Müller — @foonathan

interesting problems • Intra-register sorting • Fast linear median-of-seven filter • Fast small-kernel convolution • Faster (?) UTF-8 to UTF-32 conversion (with AVX2) • No heavy code, but lots of pictures ksize(7) ksize(9) ksize(11) ksize(13) ksize(15)Copyright © 2020 Bob Steagall UTF-8 to UTF-32 Conversion Algorithm 32 CppCon 2020 - Adventures in SIMD ThinkingCopyright © 2020 Bob Steagall K E W B C Bob Steagall K E W B C O M P U T I N G 34 CppCon 2020 - Adventures in SIMD Thinking The Basic Conversion Algorithm (UTF-8 to UTF-32) KEWB_ALIGN_FN std::ptrdiff_t UtfUtils::BasicConvert(char8_t const*

store many more numbers and allows fast access to these numbers, and to a hard disk or SSD device, which can permanently store the numbers but does not allow such fast access. The stored numbers are most process large amounts of data very quickly. The key to their enormous computing power lies in many fast CPUs working in parallel; problems or tasks are split into many small parts that are solved by individual converter to transform that program into the basic instructions that the CPU can execute. For the conversion process, we essentially have two options: we can either convert the entire program into machine

store many more numbers and allows fast access to these numbers, and to a hard disk or SSD device, which can permanently store the numbers but does not allow such fast access. The stored numbers are most process large amounts of data very quickly. The key to their enormous computing power lies in many fast CPUs working in parallel; problems or tasks are split into many small parts that are solved by individual converter to transform that program into the basic instructions that the CPU can execute. For the conversion process, we essentially have two options: we can either convert the entire program into machine

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 340 Integer to string conversion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . store many more numbers and allows fast access to these numbers, and to a hard disk or SSD device, which can permanently store the num­ bers but does not allow such fast access. The stored numbers are most process large amounts of data very quickly. The key to their enormous computing power lies in many fast CPUs working in parallel; problems or tasks are split into many small parts that are solved by individual