You're reading from Hands-On GPU Computing with Python

Product typeBook

Published inMay 2019

Reading LevelIntermediate

PublisherPackt

ISBN-139781789341072

Edition1st Edition

Languages

Python

Tools

TensorFlow CUDA

Concepts

Data Processing

Author (1)

Avimanyu Bandyopadhyay

Fundamentals of GPU Programming

In this chapter, we will be moving on from the hardware perspective of GPUs toward a computing perspective, as is the primary objective of this book. We will begin with an introduction to GPU programming and fundamental ways to set up three different platforms, namely CUDA, ROCm, and Anaconda. NVIDIA and AMD GPUs will be revisited here to explore the practical usage of GPUs with the three platforms.

The concept of Python programming integrated with GPU code invocation will be discussed. Anaconda users and Python programming enthusiasts will be motivated to invoke GPUs within their program code with CuPy and Numba (formerly Accelerate) via Anaconda. Additionally, we will learn about a few basics of hands-on GPU programming, GPU programmable platforms, CUDA, CUDA libraries, OpenCL, and ROCm.

Moving on, we will explore the Python programming aspect...

GPU-programmable platforms

Since our primary subject is GPU computing, it is essential that we learn the fundamentals of GPU programming from a computing perspective. GPU programming is the foundation of GPU computing. The question that should be pondered upon here, is why is it significant in GPU computing? To know the answer, let's first understand the basic differences between programming and computing. Here is a short comparison:

Programming	Computing
Involves developing programs.	Involves performing computations through programs.
Intends to solve multiple problems in a generalized manner.	Focused on solving a single problem that can be interdisciplinary in nature.
Programming is applied to broader scenarios.	Computing is applied to a single specific scenario that is based on and chosen from those broader scenarios.
Here is a simple example...

Basic CUDA concepts

Since we have discussed various points of Compute Unified Device Architecture (CUDA) in our earlier chapters, let's now focus on its technical aspects and GPU programmability.

Apart from the installation procedure that's unique to CUDA, the remaining concepts that will be discussed here in brief will be useful for getting started with GPU programming on all platforms.

Installing and testing

Before we discuss some fundamental GPU programming concepts, it is essential that we revisit the CUDA installation and testing procedure that we covered earlier in Chapter 2, Designing a GPU Computing Strategy, while concluding the DIY section. Step-by-step screenshots of a fresh Ubuntu 18.04 Linux installation...

Basic ROCm concepts

In Chapter 3, Setting Up a GPU Computing Platform with NVIDIA and AMD, we discussed AMD ROCm and also compared it with NVIDIA CUDA. Like we did in the previous section of this chapter, let's look into the Radeon Open Compute Platform in a similar manner.

AMD ROCm includes a set of fundamental ways to set up a GPU programming platform for Open Compute. The basic ROCm concepts you need to know to start programming on AMD or NVIDIA GPUs are as follows.

Installation procedure and testing

According to the official documentation, an ROCm installation requires performing the following Terminal-based tasks that are compatible with Ubuntu 18.04:

Update your system and install libnuma-dev using the following...

The Anaconda Python distribution for package management and deployment

Anaconda is a distribution for package management and deployment. It facilitates the development of scientific computing and machine learning through Python and R (a programming language focused on statistical computing). Anaconda simplifies the process of managing various packages and also their deployment. The Anaconda repository maintains more than 1,000 professionally built packages for data science.

It has a package management system called conda to install various scientific packages. It also provides a build feature for building your own Python packages and uploading them to Anaconda servers. conda can be used for installing, executing, and also updating packages, along with their dependencies. It can facilitate software for any language, even though it was made for Python packages.

The Anaconda distribution...

GPU-enabled Python programming

The fundamental concept behind Python programming on GPU devices is based on what we have learned so far about CUDA, ROCm, and Anaconda. It is all about using their integrations with Python developed as PyCUDA, PyOpenCL, CuPy, and Numba, respectively.

With PyCUDA, you can use Python with NVIDIA GPUs, while with PyOpenCL, you can use NVIDIA, AMD GPUs and other massively parallel compute devices. CuPy allows you to implement NumPy like features on an NVIDIA GPU. After installing Accelerate with Conda, you can import the numba package very easily within your code to leverage your GPU device. We will explore this in detail in Chapter 8, Working with Anaconda, CuPy, and Numba for GPUs.

The dual advantage

...

Summary

In this chapter, we learned about the basic differences between programming and computing. We learned about some of the fundamental concepts regarding how CUDA, ROCm, and Numba leverage GPUs. We also learned the many libraries facilitated by CUDA, ROCm, and Numba. The features of PyCUDA, PyOpenCL, and Numba were mentioned and highlighted.

Now that we're at the end of this chapter, you should be able to install CUDA, ROCm, and Anaconda on an Ubuntu-based system. You should also be able to set up the hipify tool and start porting existing CUDA code to its HIP version, especially if you are a research-code enthusiast. You are now familiar with the configurational differences between OpenCL with CUDA and OpenCL with ROCm. You have also learned the various reasons behind why Python is a great choice for GPU programming.

Before we start our hands-on experience with programming...

Official PyCUDA home page at https://mathema.tician.de/software/pycuda/
Official PyOpenCL home page at https://mathema.tician.de/software/pyopencl/
Official CuPy website at https://cupy.chainer.org/
Official Numba website at https://numba.pydata.org
Conventional CPU Computing vs GPU Computing: http://www.mediafire.com/file/dv41lvwd9klnat2/Conventional_CPU_Computing_vs_GPU_Computing.pdf

The rest of the chapter is locked

You have been reading a chapter from

Hands-On GPU Computing with Python

Published in: May 2019Publisher: PacktISBN-13: 9781789341072

A free Packt account unlocks extra newsletters, articles, discounted offers, and much more. Start advancing your knowledge today.

undefined

Unlock this book and the full library FREE for 7 days

Get unlimited access to 7000+ expert-authored eBooks and videos courses covering every tech area you can think of

Start free trial

Renews at $15.99/month. Cancel anytime

Author (1)

Avimanyu Bandyopadhyay

Avimanyu Bandyopadhyay is currently pursuing a PhD degree in Bioinformatics based on applied GPU computing in Computational Biology at Heritage Institute of Technology, Kolkata, India. Since 2014, he developed a keen interest in GPU computing, and used CUDA for his master's thesis. He has experience as a systems administrator as well, particularly on the Linux platform. Avimanyu is also a scientific writer, technology communicator, and a passionate gamer. He has published technical writing on open source computing and has actively participated in NVIDIA's GPU computing conferences since 2016. A big-time Linux fan, he strongly believes in the significance of Linux and an open source approach in scientific research. Deep learning with GPUs is his new passion!
Read more about Avimanyu Bandyopadhyay

Other recommended products

Related to this chapter

Hands-On GPU-Accelerated Computer Vision with OpenCV and CUDA

This book is a guide to explore how accelerating of computer vision applications using GPUs will help you develop algorithms that work on complex image data in real time. It will solve the problems you face while deploying these algorithms on embedded platforms with the help of development boards from NVIDIA such as the Jetson TX1, Jetson TX2, and Jetson TK1.

BookSep 2018380 pages

CUDA Cookbook

This book is for programmers who want to delve into parallel computing, become part of the high-performance computing community and apply those techniques to build modern applications. Experience with C++ programming is assumed. There are some sample examples on equivalent Fortran code. For Deep Learning enthusiasts python based sample code is also provided.

BookSep 2019508 pages

Hands-On GPU Programming with Python and CUDA

GPUs are designed for maximum throughput, but are subject to low-level subtleties. In contrast, Python is a high-level language that favours ease of use over speed. In this book, we will combine the power of both Python and CUDA to help you create high performing Python applications by using open-source libraries such as PyCUDA and SciKit-CUDA.

BookNov 2018310 pages

Python Parallel Programming Cookbook

Python Parallel Programming Cookbook, Second Edition, covers recipes that will help you how to build multithreaded, multiprocess and asynchronous applications in Python. The book will help you build applications for the GPU using CUDA and PyOPENCL and implement effective debugging and testing techniques.

BookSep 2019370 pages

Personalised recommendations for you

Based on your interests and search pattern

Et al.

Ever wonder why speech recognition systems don't understand the Scottish accent, or what would happen if an astronaut only ate mac 'n' cheese, or other spurious reflections you'd have at a bar? We did, then collated those deliberations into absurd research articles with fake figures and methodologies inspired by even more fictionally absurd studies.

BookAug 2023230 pages5

Generative AI with LangChain

This book is a comprehensive introduction to LLMs and LangChain, demystifying the basic mechanics of LangChain, its functionalities, and the myriad of applications it can be integrated into.

BookDec 2023360 pages4

Generative AI with LangChain

This book is a comprehensive introduction to LLMs and LangChain, demystifying the basic mechanics of LangChain, its functionalities, and the myriad of applications it can be integrated into.

BookDec 2023360 pages5

Generative AI with LangChain

This book is a comprehensive introduction to LLMs and LangChain, demystifying the basic mechanics of LangChain, its functionalities, and the myriad of applications it can be integrated into.

BookDec 2023360 pages1

Generative AI with LangChain

This book is a comprehensive introduction to LLMs and LangChain, demystifying the basic mechanics of LangChain, its functionalities, and the myriad of applications it can be integrated into.

BookDec 2023360 pages5

Mastering Tableau 2023

This book is a comprehensive resource to mastering your Tableau skills and becoming a BI expert. As you progress, you will learn how to build advanced dashboards and improve your storytelling to derive key business insight, as well as make you well-versed with advanced functionalities of Tableau in the business intelligence domain.

BookAug 2023684 pages

Building AI Applications with ChatGPT APIs

This guide covers all ChatGPT API features for effortless creation of robust AI powered apps. With its help, you’ll be able to leverage ChatGPT’s cutting-edge NLP models to take your app development skills to the next level. You’ll also work on ten exciting projects that will give you the practical know-how that you can apply to your existing applications.

BookSep 2023258 pages5

Building AI Applications with ChatGPT APIs

This guide covers all ChatGPT API features for effortless creation of robust AI powered apps. With its help, you’ll be able to leverage ChatGPT’s cutting-edge NLP models to take your app development skills to the next level. You’ll also work on ten exciting projects that will give you the practical know-how that you can apply to your existing applications.

BookSep 2023258 pages2

Data Engineering with AWS

Embark on a journey to master data engineering pipelines on AWS! Our book offers a hands-on experience of AWS services for ingesting, transforming, and consuming data. Whether you're an absolute beginner or someone with basic data engineering experience, this guide is an indispensable resource.

BookOct 2023636 pages5

Modern Data Architecture on AWS

Every organization wants an agile, performant, and cost-effective data platform that meets all their current and future business needs. Purpose-built AWS analytics services and their features play a big part in building such a modern data platform. This book brings to you all the design and architectural patterns that’ll help you achieve this goal.

BookAug 2023420 pages5

Practical Guide to Applied Conformal Prediction in Python

Discover the power of Conformal Prediction with the "Practical Guide to Applied Conformal Prediction in Python." Master the latest techniques to quantify uncertainty in machine learning and computer vision models, and seamlessly apply them to your industry applications.

BookDec 2023240 pages

TinyML Cookbook

With over 70 project-based recipes, the TinyML Cookbook is a practical guide that will help you to get the most out of your microcontrollers. It provides a comprehensive understanding of the theoretical foundations while giving you hands-on experience training ML models for deployment on Arduino Nano 33 BLE Sense, Raspberry Pi Pico, and SparkFun RedBoard Artemis Nano microcontrollers.

BookNov 2023664 pages