Programming Persistent Memory

A Comprehensive Guide for Developers

 

Read or Download

Explore

Step 1: Learn About Persistent Memory Uses

Explore how this technology can address challenges with use case memory.

Learn More

Step 2: Analyze Your Application

Use the Platform Profiler for Intel® VTune™ Profiler to learn how your application can benefit from large capacity persistent or volatile memory.

Try Platform Profiler

Free Download of Intel VTune Profiler

Step 3: Configure Your Platform

Provision Intel Optane DC persistent memory.

Get Started

Quick Start Guide

Provision Persistent Memory in Linux*

Step 4: Develop Software with the Persistent Memory Development Kit

Use this set of libraries and tools to begin developing persistent memory solutions.

Get Started

Get Started

Introduction to Persistent Memory Programming

Use the Persistent Memory Development Kit (PMDK) to create a new application that takes advantage of Intel® Optane™ DC persistent memory.

Browse the Library

 

Operating Modes

Intel Optane DC persistent memory modules support two operating modes:

  • Memory Mode enables an application to operate on a large pool of volatile memory without modification.
  • App Direct Mode offers a low-latency, byte-addressable pool of persistent memory.

Use the information here to help modify or develop an application with low latency and data persistence.

Key Concepts

Programming to byte-addressable persistent memory requires new ways of thinking about data persistence and consistency.

What You Need to Know

About the PMDK

The PMDK is an open source collection of libraries and tools designed to simplify development, debugging, and management of persistent memory-aware applications.

PMDK Overview

Configure Development Environments

Linux*

Options include emulation or a persistent memory-aware hypervisor, such as QEMU or VMware vSphere* virtual machine. 

Configure Intel Optane DC Persistent Memory

Details

Windows*

Choose from several support options for Windows Server 2019. Access guidance from Microsoft* for Windows Server* 2016.

Details

Set Up the PMDK

Download binaries or build from the latest source. The PMDK is sent to many Linux* distribution package repositories.

Linux

Install Using Linux Packages

Install from Source Code

Windows

Install Using vcpkg

User Guides

ndctl

ipmctl

Quick Start Configuration Guide

Get Started with PMDK

Evaluate

Analyze the entire system to identify configuration issues and workloads that can benefit from persistent memory or analyze individual applications to determine optimization opportunities.

System View: Intel® VTune™ Profiler—Platform Profiler

Application View: Intel® VTune™ Profiler—Memory Analysis

Design

Plan for persistence with these step-by-step guides that illustrate programming concepts.

A Simple grep Example

A Fault-Tolerant Algorithm for MapReduce

Create a MySQL* Storage Engine

Get High-Level Language Support

Learn PMDK basics with introductory tutorials and code samples.

C

Code Samples

C++

Hello World

Code Samples

Java*

Hello World for Persistent Collections for Java* (PCJ)

Hello World for Low Level Persistence Library (LLPL)

Java API for Persistent Memory Programming

Code Samples

Debug & Troubleshooting Tools

Tools from Intel

Find persistence errors quickly with Intel® Inspector—Persistence Inspector.

Detect Programming Errors with Cache Management

Find Out-of-Order Stores in Your Persistent Memory Program

PMDK Tools

Pmemcheck

Perform dynamic run-time analysis with an enhanced version of Valgrind*.

 

Pmempool

Manage and analyze persistent memory pools with this stand-alone utility.

Tools

Intel® VTune™ Profiler

This premier performance profiler has new capabilities to help you optimize your persistent memory programs.

Free Download

Platform Profiler

Analyze systems over longer intervals. Find out which workloads can benefit from larger memory allocations and which system configuration better fits the workloads.

Memory Access Profiling

Locate code that is sensitive to memory bandwidth and latency issues. Identify hot, warm, or cool data to optimize memory usage and placement.

I/O Profiling

Identify opportunities to replace disk or SSD-based storage with faster persistent memory.

Intel® Inspector—Persistence Inspector

This tool finds persistence errors quickly to make software fast and reliable. It checks that all caches only flush once to persistent memory and written in the correct order.

How to Detect Errors

Free Download

Open Source Tools

Manage persistent memory modules, check for programming errors, and evaluate performance.

ndctl

Manage the Linux* LIBNVDIMM kernel subsystem.

pmemcheck

Perform a dynamic runtime analysis with an enhanced version of Valgrind*.

pmempool

Manage and analyze persistent memory pools with this stand-alone utility.

FIO

Run benchmarks with FIO.

pmembench

Build and run PMDK benchmarks.

Training

Persistent Memory Development Kit (PMDK)

This open source kit includes a set of libraries and tools to support software development for this new technology.

Boost Your C++ Applications with Persistent Memory

This code sample demonstrates using the C++ bindings of libpmemobj to convert a simplified version of grep, the Unix* command-line utility, to use persistent memory.

Implement a Fault-Tolerant Algorithm

This code sample uses libpmemobj C++ bindings to demonstrate how to implement fault tolerance in a PMEM version of the famous MapReduce algorithm.

Introduction to Persistent Collections for Java*

Learn how to instantiate, store, and fetch persistent data after a power cycle.

Create a Persistent Memory-Aware Queue

Make queue operations transactional to prevent persistent memory corruption. Core concepts are demonstrated in sample code.

Panaconda: A Persistent Memory Version of the Game Snake

This code sample demonstrates PMDK APIs for pools, pointers, and transactions, as well as shows you how to build and run the game.

PMAN: A Persistent Memory Version of the Game Pac-Man

The PMAN code sample highlights program design, persistent memory pools, pointers, and transactions. Like Panaconda, you can run the example.

Give Feedback