LifeIsStrange

High-Performance Matrix Multiplication

This is a short post that explains how to write a high-performance matrix multiplication program on modern processors. In this tutorial I will use a single core of the Skylake-client CPU with AVX2, but the principles in this post also apply to other processors with different instruction sets (such as AVX512).

Intro

Matrix multiplication is a mathematical operation that defines the product of

Software Versions

$ runc --version
runc version 1.0.0-rc4+dev
commit: f53ad9cec928821a25c2a74eb8bd95558f8cf152-dirty
spec: 1.0.0

$ containerd --version
containerd github.com/containerd/containerd v1.0.0-beta.3-86-g9649a42 9649a428e8c470cddc0d5a3c394f36111f940adc

Thread Pools

Thread pools on the JVM should usually be divided into the following three categories:

CPU-bound
Blocking IO
Non-blocking IO polling

Each of these categories has a different optimal configuration and usage pattern.

I recently happened upon a very interesting implementation of popen() (different API, same idea) called popen-noshell using clone(2), and so I opened an issue requesting use of vfork(2) or posix_spawn() for portability. It turns out that on Linux there's an important advantage to using clone(2). I think I should capture the things I wrote there in a better place. A gist, a blog, whatever.

This is not a paper. I assume reader familiarity with fork() in particular and Unix in general, though, of course, I link to relevant wiki pages, so if the unfamiliar reader is willing to go down the rabbit hole, they should be able to come ou

	import spacy
	from spacy.matcher import Matcher

	Active = "Harry ate six shrimp at dinner.\
	Beautiful giraffes roam the savannah.\
	Sue changed the flat tire.\
	We are going to watch a movie tonight.\
	I ran the obstacle course in record time.\
	The crew paved the entire stretch of highway.\
	Mom read the novel in one day.\

	%% Topic : Binary Search Tree Implementation in Prolog
	%% Author : Abhijeet Sharma \| Akhil Singhal
	%% Date : 26/03/2018
	%% Course : Paradigms of Programming

	%% Sample Inputs

	%% Input inl([3,6,1,2,4,7,9],nil,X).
	%% X = t(3, t(1, nil, t(2, nil, nil)), t(6, t(4, nil, nil), t(7, nil, t(9, nil, nil))))

	##
	# This is an adoption from Oh My Zsh for Google Cloud Shell
	# author - bobvanluijt
	# source - https://github.com/bobvanluijt/google-cloud-shell-pro-zsh
	##

	main() {

	# Install zsh
	sudo apt-get -qq update

	# If you come from bash you might have to change your $PATH.
	export PATH=$HOME/bin:/usr/local/bin:$PATH

	# Path to your oh-my-zsh installation.
	export ZSH=$HOME/.oh-my-zsh

	# Set name of the theme to load. Optionally, if you set this to "random"
	# it'll load a random theme each time that oh-my-zsh is loaded.
	# See https://github.com/robbyrussell/oh-my-zsh/wiki/Themes
	ZSH_THEME="agnoster"

	import kotlin.coroutines.experimental.*
	import kotlin.coroutines.experimental.intrinsics.*

	fun main(args: Array<String>) {
	enumerate {
	if (flip("A")) {
	if (flip("B")) 1 else 2
	} else {
	if (flip("C")) 3 else if (flip("D")) 4 else 5
	}