Optimizing Floating Point Math in Julia | Oscar Smith | JuliaCon 2022

preview_player
Показать описание
Why did `exp10` get 2x faster in Julia 1.6? One reason is, unlike most other languages, Julia doesn't use the operating system-provided implementations for math (Libm). This talk will be an overview of improvements in Julia's math library since version 1.5, and areas for future improvements. We will cover will be computing optimal polynomials, table based implementations, and bit-hacking for peak performance.

00:00 Welcome!
00:10 Help us add time stamps or captions to this video! See the description for details.

  1. The Julia Programming Language
Рекомендации по теме
Optimizing Floating Point 0:29:41

Optimizing Floating Point Math in Julia | Oscar Smith | JuliaCon 2022

Building a fully 0:09:20

Building a fully verified, optimizing compiler for floating-point arithmetic

An Introduction To 0:45:24

An Introduction To Floating Point Math - Martin Hořeňovský - NDC TechTown 2022

Wait, so comparisons 0:15:41

Wait, so comparisons in floating point only just KINDA work? What DOES work?

Webinar 034: Tools 1:06:54

Webinar 034: Tools and Techniques for Floating-Point Analysis

[POPL'18] On Automatically 0:25:30

[POPL'18] On Automatically Proving the Correctness of math.h Implementations

Tech talk: Custom-Precision 0:33:07

Tech talk: Custom-Precision Mathematical Library Explorations for Code Profiling and Optimization

Floating Point Bit 0:19:31

Floating Point Bit Hacks Every Programmer Should Know (Including Fast Inverse Square Root - Quake)

Portable floating-point calculations 0:58:22

Portable floating-point calculations - Guy Davidson - Meeting C++ 2024

L25 Floating Point 0:17:08

L25 Floating Point

Why Is This 0:05:46

Why Is This Happening?! Floating Point Approximation

a compiler flag 0:01:00

a compiler flag that enables a set of aggressive floating-point optimizations.

Understanding Floating Point 0:06:13

Understanding Floating Point Math and Errors

C++ : C++, 0:01:19

C++ : C++, How floating-point arithmetic operations get optimized?

Floating-point Arithmetic | 0:40:41

Floating-point Arithmetic | Week 10 | MIT 18.S191 | Alan Edelman

FP8 Doing Math 0:10:57

FP8 Doing Math with Floating Point Numbers

Understanding the Differences 0:01:40

Understanding the Differences in Floating Point Values Between Release and Debug Builds

2019 LLVM Developers’ 0:21:31

2019 LLVM Developers’ Meeting: J. Garzella “Floating Point Consistency in the Wild: ”

Optimized Solution for 0:20:00

Optimized Solution for Floating Point Computing, Nimrod Alexandron, Cadence (ChipEx2021)

y=mx+c explained | 0:00:25

y=mx+c explained | GCSE Maths #shorts

What Is Half 0:02:15

What Is Half Precision?

computers suck at 0:05:09

computers suck at division (a painful discovery)

[SAS23] Modular Optimization-Based 0:22:26

[SAS23] Modular Optimization-Based Roundoff Error Analysis of Floating-Point Programs

C++ : Floating-point 0:01:33

C++ : Floating-point optimizations - guideline

Комментарии
Автор

Great talk, really informative and super helpful

edbed
Автор

So can the user choose the floating point precision when starting Julia ? (if sm want to try to use 3ulp precision for exemple)

BigToinE