According to new data from the IDC Worldwide Semiannual Public Cloud Services Tracker, global revenue for the public cloud services market will total $545.8 billion in 2022, up 22.9% from 2021.
SaaS applications are the largest revenue generator for public cloud services, accounting for over 45% of the total by 2022. Infrastructure as a Service was the second largest revenue category at 21.2% of total revenue, followed by Platform as a Service and SaaS System Infrastructure. Software was 17.0% and 16.7% respectively.
According to IDC data, the top five public cloud providers (Microsoft, Amazon Web Services, Salesforce Inc., Google, Oracle) account for more than 41% of total global revenue, growing 27.3% year over year. shown. In the public cloud services market, Microsoft held the largest market share with 16.8%, followed by Amazon Web Services with 13.5%.
“Given the economic challenges of the past year, a focus on containing new spending and optimizing the use of existing cloud assets will dominate CIO priorities and determine the next destiny of IT providers. It’s easy to conclude that the time has come, a few years, and that’s also a very false conclusion,” said Rick Villers, group vice president of Worldwide Research at IDC. “Evaluation and use of AI, in the wake of generative AI, is starting to take center stage in planning, and the long-term investment challenges of enterprises and cloud providers will play an important role in evaluating and deploying AI enablement services. increase.”
Generative AI Drives Cloud Investments
Prolonged economic uncertainty may be straining some budgets, but generative AI is driving long-term investment strategies for enterprises and cloud providers.
Dave McCarthy, research vice president of cloud and edge infrastructure services at IDC, said: “This serves two purposes. First, it opens up the next wave of migration for enterprise applications that have traditionally stayed on-premises. Second, it creates a foundation for new AI software that can be rapidly deployed at scale. In both cases, these investments present opportunities for market growth.”
Venture capital firm Andreessen Horowitz says generative AI is resource-intensive in nature, requiring significant cloud investments.
“Nearly everything in generative AI goes through a GPU (or TPU) hosted in the cloud at some point. A model provider/research institute running a training workload, a hosting company doing inference/fine-tuning, FLOPS is the lifeblood of generative AI, whether you’re an application company doing a mix of the two, or a mix of both,” the company wrote. “For the first time in a very long time, the most disruptive advances in computing technology depend on large-scale computing.”
Andreessen Horowitz says that access to computing resources at the lowest total cost is a determinant of success for AI companies. Venture capital firms expect the majority of start-ups to use cloud computing for generative AI, often due to low upfront costs and high scalability.
Cloud providers are striving to keep up with AI demand
Recent wall street journal The article pointed out that traditional cloud infrastructure was not designed to support AI at scale, and that cloud providers are rushing to keep up with demand.
“There is a pretty big imbalance between supply and demand right now,” said Chetan Kapoor, director of product management for Amazon Web Services’ Elastic Computing Cloud division. WSJMore article.
Only a few cloud services are optimized for AI. While the majority of the cloud consists of servers utilizing general-purpose CPUs, GPU clusters suitable for running AI workloads make up a minority of the infrastructure.
Kapoor said. WSJMore AWS plans to deploy multiple AI-optimized server clusters over the next 12 months. The article also mentioned that Microsoft Azure and Google Cloud are also working to strengthen their AI infrastructure.
Hewlett Packard Enterprise is also entering the AI cloud market. The company recently announced that it will introduce HPE GreenLake for Large Language Models, an on-demand, multi-tenant supercomputing he cloud service that enables enterprises to privately train, tune and deploy AI at scale. .
“Unlike general-purpose cloud offerings that run multiple workloads in parallel, HPE GreenLake for LLM is uniquely designed to run a single large-scale AI training and simulation workload at full computing power. It runs on an AI-native architecture that has the power to do the job,” the company said. on release. “This product supports AI and HPC jobs on hundreds or thousands of CPUs or GPUs at once. This feature is much more effective in training AI and creating more accurate models. , is reliable and efficient, enabling companies to move from POC to production faster and resolve issues faster.”
HPE President and CEO Antonio Neri commented that the AI market has reached a generational shift as transformative as previous technological breakthroughs such as web, mobile and cloud.
“HPE is helping what was once a well-funded government lab and a global Once a big field, AI is democratized, and now organizations can take advantage of AI with on-demand cloud services to train, tune, and deploy models responsibly at scale. , can drive innovation, disrupt markets and achieve breakthroughs,” he said.
This story was originally published on our sister site, Enterprise AI.
