[Paper Review] Cloud Platforms for Developing Generative AI Solutions: A Scoping Review of Tools and Services

Dhavalkumar D. Patel, Ganesh Raut+6 more|arXiv (Cornell University)|Dec 8, 2024

Big Data and Business Intelligence5 citations

TL;DR

A scoping review comparing major cloud providers and their tools for developing generative AI, covering compute, data, security, edge, and governance aspects, with case studies and future directions.

ABSTRACT

Generative AI is transforming enterprise application development by enabling machines to create content, code, and designs. These models, however, demand substantial computational power and data management. Cloud computing addresses these needs by offering infrastructure to train, deploy, and scale generative AI models. This review examines cloud services for generative AI, focusing on key providers like Amazon Web Services (AWS), Microsoft Azure, Google Cloud, IBM Cloud, Oracle Cloud, and Alibaba Cloud. It compares their strengths, weaknesses, and impact on enterprise growth. We explore the role of high-performance computing (HPC), serverless architectures, edge computing, and storage in supporting generative AI. We also highlight the significance of data management, networking, and AI-specific tools in building and deploying these models. Additionally, the review addresses security concerns, including data privacy, compliance, and AI model protection. It assesses the performance and cost efficiency of various cloud providers and presents case studies from healthcare, finance, and entertainment. We conclude by discussing challenges and future directions, such as technical hurdles, vendor lock-in, sustainability, and regulatory issues. Put together, this work can serve as a guide for practitioners and researchers looking to adopt cloud-based generative AI solutions, serving as a valuable guide to navigating the intricacies of this evolving field.

Motivation & Objective

• Assess state-of-the-art cloud services for generative AI from major providers (AWS, Azure, GCP, IBM, Oracle, Alibaba) and their strengths/weaknesses.
• Evaluate capabilities for large language models, multimodal AI, HPC, serverless, edge computing, and data management.
• Identify security, ethical, and regulatory challenges and opportunities for innovation in cloud-based generative AI.
• Provide practical guidance and strategic recommendations for practitioners, researchers, and policymakers.

Proposed method

• Systematic scoping review following Arksey and O'Malley framework.
• Comprehensive literature and provider documentation search across academic databases and vendor materials.
• Comparative framework assessing compute, AI services, data management, pricing, and emerging technologies.
• Consultation with industry experts and academics to validate findings.
• Synthesis includes SWOT analyses and future directions in areas like edge AI, federated learning, and sustainability.

Figure 1: Structured guide through cloud-based generative AI development landscape

Experimental results

Research questions

• RQ1 What are the capabilities and limitations of major cloud platforms for generative AI development?
• RQ2 How do providers compare in terms of HPC, AI services, data management, and deployment tooling for generative AI?
• RQ3 What security, ethical, regulatory, and sustainability challenges accompany cloud-based generative AI?
• RQ4 What are the emerging trends and future directions in cloud-enabled generative AI development?

Key findings

• AWS offers a robust Generative AI Stack with SageMaker and Bedrock for foundation models and GPUs like P4d for large-scale training.
• Azure emphasizes enterprise integration, OpenAI collaboration, and responsible AI tooling within its Generative AI Stack.
• GCP highlights Vertex AI, Cloud TPUs, and multimodal capabilities with initiatives like MUM.
• Providers emphasize end-to-end AI lifecycles, data/AI integration, and governance through explainability and bias tools.
• The cloud AI market is dominated by AWS, Azure, and GCP, collectively shaping AI services with significant market shares.
• There is growing adoption of hybrid, multi-cloud, and edge computing strategies to optimize AI deployment and data governance.

Figure 2: Major Milestone Timeline in AI Development