The twin pillars of AI and cloud computing together are propelling businesses forward. Worldwide spending on public cloud services is forecast to swell from $679 billion in 2024 to $1 trillion in 2027, while the AI market will grow from $184 billion in 2024 to $407 billion in that same period.
Recent developments that suggest that AI and cloud computing are marching forward in lock step include:
Microsoft has integrated OpenAI's models, including ChatGPT, into its Azure cloud platform, and telecommunications leaders like Windstream, AudioCodes, AT&T, and Vodafone are already leveraging the Microsoft Azure OpenAI service to better engage with their customers and streamline their operations.
In June 2024, Oracle announced a partnership with OpenAI to extend the Microsoft Azure AI platform to Oracle Cloud Infrastructure (OCI).
Amazon Web Services is delivering generative AI capabilities, with users leveraging enterprise-grade security and privacy to scale their specific use cases.
Given this emerging intersection of cloud and AI, enterprise leaders should strive to understand how the rapidly expanding field of AI works in tandem with cloud computing technology to spur ever-greater innovation.
Why Cloud Computing and AI Make a Great Team
Businesses can utilize their existing cloud-skilled professionals to help drive AI adoption. That’s because AI requires not just new applications, but “good infrastructure people along with the developers who build the apps,” Nutanix President and CEO Rajiv Ramaswami told The Forecast.
Much of today’s infrastructure talent has been trained in cloud computing, and those skills, in turn, can be leveraged to accelerate AI.
There’s also symbiosis between AI and cloud computing in the realm of automation. Implementing AI can streamline simple processes, increasing efficiency and allowing IT talent to focus more on innovative development.
Cloud can help to overcome the challenge posed by AI’s rising demands for compute power.
“It must be simultaneously exciting and terrifying to be a data center manager right now,” Greg Diamos, co-founder of AI company Lamini, told The Forecast.
“You don't have enough compute in your data center, no matter who you are.”
Cloud delivers the massive compute capability needed to keep up with the demands of AI.
“Overall, the two technologies naturally complement one another, particularly when one looks at the potential within multi-cloud environments,” said Induprakas Keri, senior vice president and general manager of hybrid multicloud at Nutanix.
While public cloud will power much of AI, organizations will still want to manage their private data using private cloud or their own data centers. That drives toward a hybrid multi-cloud IT model.
“AI-based services and applications are absolutely made for hybrid multi-cloud architectures,” Keri said.
“Steps in the AI workflow will happen across various infrastructure environments, with training happening in the cloud, enrichment, refinement, and training in core data centers, and inferencing at the edge. Successfully delivering a cohesive, scale-out infrastructure that can span across this entire AI workflow will be a key to success.”
Many are already seizing upon the opportunity here. A Deloitte study, for example, found that 70% of companies get their AI capabilities through cloud-based software, while 65% create AI applications using cloud services. There’s a natural fit here. Most AI apps are built with cloud native (Kubernetes) technology to move across IT infrastructures.
“The cloud has turned out to be an amazing distribution mechanism for algorithms – all three of the leading cloud providers have made available a set of algorithms that make AI much easier to do,” explained Tom Davenport, President's Distinguished Professor of IT & Management at Babson College.
Where AI and Cloud Computing Converge
AI and cloud computing converge in automating data analysis, data management, security, and decision-making processes. The ability of AI to exercise machine learning and to derive impartial interpretations of data-driven insights fuels efficiency in these processes and can lead to significant cost savings on numerous fronts within the enterprise.
Using AI software based on machine learning algorithms in cloud environments delivers intuitive and connected experiences for customers and users. In the consumer world, Alexa and Siri are two examples of this seamless combination that enables various operations, from conducting a search to playing a song to making a purchase.
Machine learning (ML) models use large data sets to train the algorithm. This data can be structured, unstructured, or raw and needs powerful CPUs and GPUs to process. Ideally, a combination of public and private or hybrid cloud systems can provide such huge amounts of compute power today. Further, AI cloud computing also enables services that are used in ML, such as serverless computing, batch processing, and container orchestration.
5 Ways AI and Cloud Computing Work Together
In general, businesses need to ponder the impact of AI. “First, they need to ask things like: Is it even worth doing? Will it help the organization build value?” said Eric Siegel, bestselling author of The AI Playbook: Mastering the Rare Art of Machine Learning Deployment. “You need to have a very specific notion of how this technology is going to improve some existing process, hopefully making it more efficient and effective.”
Likewise, organizations shouldn’t be looking at the convergence of cloud and AI for its own sake. Rather, they should be weighing the practical benefits.
With public cloud services, developers do not need to build and manage a separate infrastructure for hosting AI platforms. They can use ready configurations and models to test and deploy AI applications.
Further, generic services based on AI but not necessarily requiring an ML model – such as speech-to-text, analytics, and visualization – can be improved by running them from the cloud using first-party data generated by the organization.
Some of the more common AI-based applications in the cloud include:
IoT – Cloud architectures and services that power the Internet of Things can store and process data generated by AI platforms on IoT devices.
Chatbots – Chatbots are ubiquitous AI-based software that use natural language processing to communicate with users. Telecommunications giant Vodafone, for example, has developed TOBi, a digital assistant supported by Azure. With chatbots, cloud platforms store and process the data, and cloud services connect them to the appropriate applications for further processing. Customer data is also fed back into the chatbot application that resides in the cloud.
Business Intelligence – Business intelligence is another mainstream application where AI cloud computing can gather data on the market, target audience, and competitors of customers. Windstream streamlined workflows and improved service quality by analyzing customer interactions using AI-powered analytics. The cloud facilitates data storage and transfer, while the AI runs it through predictive analytics models.
AI as a Service (AIaaS) – Public cloud vendors now offer AI outsourcing services, allowing companies to test out software and ML algorithms without risking their primary infrastructure. They can deploy off-the-shelf AI applications at a fraction of the cost of in-house AI with significant CAPEX savings.
Cognitive cloud computing – Cognitive computing uses AI models to replicate and simulate human thought processes in complex situations. Players such as IBM and Google have built cognitive cloud platforms that provide cognitive insights-as-a-service to enterprises and facilitate the application of this technology in finance, retail, healthcare, and other industries.
Why Deploy AI in Cloud Environments – 5 Reasons
AI is the cherry on the cloud cake and the frosting, ganache, strawberries, and sprinkles combined! Here’s why AI and cloud computing form a winning team.
Cost Savings – Traditionally, ML-based models ran on expensive machines with multiple GPUs in enterprise data centers. The ever-expanding demands of AI ask for more: Businesses need to scale up GPU, CPU, and other computing resources.
“I need to figure out how to build the infrastructure to support that because traditional compute does not account for it,” said Steve McDowell, chief analyst at NAND Research.
With advances in virtualization in both public and private clouds, the cost of building, testing, and deploying these models has decreased drastically. This has leveled the playing field for many small-to-medium businesses.
“Back when I first got out of college, we were building things that literally cost $100 million of data center space just to get simple questions answered,” said David Linthicum, Chief Cloud Strategy Officer at Deloitte Consulting. “I can start up my AI skillsets with just a credit card these days.”
Productivity – AI-based algorithms required significant admin time and effort in terms of building testing and production environments, software management, and provisioning hardware resources for compute operations and storage. A centrally managed hybrid cloud or a public cloud does away with this, leaving IT staff to focus on non-repetitive tasks.
Automation – AI cloud computing is also being embedded right into the infrastructure to help automate routine processes and streamline workloads. In a hybrid cloud environment, AI tools can be used to monitor, manage, and self-heal individual public and private cloud components. AnalyticsData residing in most cloud workloads needs to be analyzed for more insights. AI-based models make it easy to mine this data in real time and develop native analytics and dashboards for each of these applications.
Data Management – AI helps boost cloud workloads in customer service, marketing, Enterprise Resource Planning, and supply chain management by processing and generating data in real time. For example, AI tools embedded in Dataflow, the streaming analytics platform in Google Cloud, can enable functions as varied as programmatic bidding in media advertising, fraud prevention in financial services, threat detection in IT security, and personalized shopping recommendations in retail.
Better SaaS Tools – Perhaps the most obvious and popular use of algorithms in AI cloud computing is their integration in mainstream SaaS tools to help these deliver more functionality and value to end users.
Challenges in Deploying AI in Cloud Environments – Merging AI and the cloud isn’t always cakes and ale. The main concerns are data privacy and connectivity.
Data Privacy – The pay-as-you-go model of SaaS technology allows thousands of companies across the world to make sense of data, find efficiencies in routine processes, develop new products, and even expand into new verticals. Therefore, they run their customer, vendor, and market data through cloud applications with little-to-no appreciation of the security risks of the public cloud. When AI processes data fed into a SaaS tool in a public cloud environment, it amplifies these risks on an exponential scale. Sensitive company data could be exposed to a security breach or unauthorized access when the processes and perimeters for AI algorithms are not clearly defined.
Connectivity – Any algorithm or data processing system in the cloud depends on one thing to keep it running: a steady internet connection. Poor network connectivity can slow down ML processes and defeat the purpose where real-time transactions and analytics are involved.
Is Hybrid Cloud the New Home of AI?
The enterprise is looking to AI for more and more real-time insights that drive innovation and give it a competitive advantage. For this, it needs a robust infrastructure that can handle vast amounts of data while guaranteeing security and functionality for end users.
Nutanix has partnered with NVIDIA and Mellanox Technologies to create an AI-ready hybrid cloud infrastructure that enables companies in retail, healthcare, finance, aerospace, and other industries to develop turnkey AI-based solutions and applications. Indeed, AI has finally found its castle in the cloud!
This is an update of an original article published June 19, 2020 and was revised on July 23, 2023.
Feature image 1 by Pixabay.
Michael Brenner has written hundreds of articles on sites such as Forbes, Entrepreneur Magazine, and The Guardian. Follow him @BrennerMichael.
© 2024 Nutanix, Inc. All rights reserved. For additional information and important legal disclaimers, please go here.