Paddy Horan paddyhoran

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

	#!/usr/bin/env python3
	import socket
	import json

	HOST, PORT = "localhost", 9527
	data = {
	"name": "hello, I am Tom.",
	"age": 10,
	"info": "sample is simple."
	}

	##########
	# Tweaked Win10 Initial Setup Script
	# Primary Author: Disassembler <[email protected]>
	# Modified by: alirobe <[email protected]> based on my personal preferences.
	# Version: 2.20.2, 2018-09-14
	# Primary Author Source: https://github.com/Disassembler0/Win10-Initial-Setup-Script
	# Tweaked Source: https://gist.github.com/alirobe/7f3b34ad89a159e6daa1/
	# Tweak difference:
	#
	# @alirobe's version is a subset focused on safely disabling telemetry, some 'smart' features and 3rd party bloat ...

	extern crate arrow;
	extern crate datafusion;

	use std::sync::Arc;
	use arrow::datatypes::{Schema, Field, DataType};
	use datafusion::execution::context::ExecutionContext;

	/// This example demonstrates executing a simple query against an Arrow data source (Parquet) and
	/// fetching results
	fn main() {

	# Selecting Columns
	df[['column1', 'column2']]

	# Filtering
	df[df.column_name == value]
	df[(df.column_name >= value) & (df.column_name < value)]
	df[(df.column_name < value) \| (df.column_name != value)]

	# Sorting
	series.order() # by data

	import sys
	reload(sys)
	sys.setdefaultencoding('utf-8')