GPU-Accelerated Deep Learning

Explore the convergence of deep learning and GPU technology. This book is a complete guide for those wishing to use GPUs to accelerate AI workflows.

The book is meant to make complex concepts understandable, with step-by-step instructions on how to set up and use GPUs in deep learning applications. Starting with an introduction to the fundamentals, you'll dive into progressive topics like Convolutional Neural Networks (CNNs) and sequence models, exploring how GPU optimization boosts performance. Further, you will learn the power of generative models, and take your skills by deploying AI models on edge devices. Finally, you will master the art of scaling and distributed training to handle large datasets and complex tasks efficiently.

This book is your roadmap to becoming proficient in deep learning and harnessing the full potential of GPUs.

What You Will Learn:

• How to apply deep learning techniques on GPUs to solve challenging AI problems.
• Optimizing neural networks for faster training and inference on GPUs
• Integration of GPUs with Microsoft Copilots

• Implementing VAEs (Variational Autoencoders) with TensorFlow and PyTorch
  Who This Book Is For:

  Industry IT professionals in AI. Students pursuing undergraduate and postgraduate degrees in Engineering, Computer Science, Data Science.

  Implementation of deep learning using GPU acceleration in Azure Covers deep learning methodologies using popular GPU-accelerated frameworks such as TensorFlow and PyTorch Discusses optimizing neural networks for faster training and inference on GPU

  Autorentext

Dr. Ramchandra Sharad Mangrulkar is a Professor in the Department of Information Technology at Dwarkadas J. Sanghvi College of Engineering in Mumbai, India. He holds various memberships in professional organizations such as IEEE, ISTE, ACM, and IACSIT. He completed his Doctor of Philosophy (Ph.D.) in Computer Science and Engineering from S.G.B. Amravati University in Maharashtra, and Master of Technology (MTech) degree in Computer Science and Engineering from the National Institute of Technology, Rourkela. Dr. Mangrulkar is proficient in several technologies and tools, including Microsoft's Power BI, Power Automate, Power Query, Power Virtual Agents, Google's Dialog Flow, and Overleaf. With over 23 years of combined teaching and administrative experience, Dr. Mangrulkar has established himself as a knowledgeable and skilled professional in his field. He has also obtained certifications such as Certified Network Security Specialist (ICSI - CNSS) from ICSI, UK. Dr. Mangrulkar has a strong publication record with 95 publications including refereed/peer-reviewed international journal publications, book chapters with international publishers (including Scopus indexed ones), and international conference publications.

Dr. Pallavi Vijay Chavan is an Associate Professor in the Department of Information Technology at Ramrao Adik Institute of Technology, D Y Patil Deemed to be University, Navi Mumbai, MH, India. She has been in academics since the past 17 years and has worked in the areas of computing theory, data science, and network security. In her academic journey, she has published research work in the data science and security domains with reputed publishers including Springer, Elsevier, CRC Press, and Inderscience. She has published 2 books, 7+ book chapters, 10+ international journal papers and 30+ international conference papers. She is currently guiding 5 Ph.D. research scholars. She completed her Ph.D. from Rashtrasant Tukadoji Maharaj Nagpur University, Nagpur, MH, India in 2017. She secured the first merit position in Nagpur University for the degree of B.E. in Computer Engineering in 2003. She is recipient of research grants from UGC, CSIR, and University of Mumbai. She is an active reviewer for Elsevier and Inderscience journals. Her firm belief is "Teaching is a mission.

Inhalt

1 Introduction to Deep Learning and GPU Acceleration.- 2 Convolutional Neural Networks (CNNs) with GPU Optimization.- 3 Sequence Models and Recurrent Networks.- 4 Generative Models and integration with Microsoft Copilots.- 5 Deployment on Edge Devices.- 6 Scaling and Distributed Training.