Well, the cloud is alive, well, and riding high on the AI wave.
Worldwide spending on public cloud services is forecast to reach $805 billion in 2024 and double in size by 2028, according to IDC, with a five-year compound annual growth rate of 19.4%.
“The continued growth we expect to see in public cloud spending can be largely attributed to gen AI due to the continued creation of general-purpose foundation models and the ramp up to delivering gen AI–enabled applications at scale,” says Gartner analyst Sid Nag. “Because of this continued growth, we expect public cloud end-user spending to eclipse the $1 trillion mark before the end of this decade.”
Public cloud market share remains relatively unchanged, with AWS the undisputed leader (32%), Microsoft Azure solidly in second (23%), and Google Cloud Platform a distant third (12%).
Combined, the Big 3 account for 67% of the total market, with smaller players, including Alibaba, IBM, and Salesforce, each in the 2-4% range.
Despite this near constancy, the emergence of generative AI has provided a market opportunity for upstarts. While the Big 3 scramble to add AI capacity to their legacy data centers, startups are building GPU-based, high-performance data centers from the ground up, offering developers the opportunity to launch and scale GPU clusters for AI training and inferencing.
This means IT organizations can avoid vendor lock-in, maintain strategic relationships with their hyperscalers of choice, but also branch out with targeted AI projects from startups who claim their purpose-built platforms are significantly faster and less expensive than incumbents.
The number of large data centers operated by hyperscalers passed the 1,000 mark in early 2024, twice the number from just four years ago. Synergy Research Group forecasts that total hyperscale data center capacity will double again in the next four years.
“Capacity growth will be driven increasingly by the even larger scale of those newly opened data centers, with generative AI technology being a prime reason for that increased scale,” Synergy Research writes.
In the first half of 2024 alone, AWS has announced $50 billion in data center investments, including $35 billion at multiple sites in Virginia, $11 billion in Indiana, and $10 billion in Mississippi.
And that’s not to mention recent announcements that see the Big 3 interested in going nuclear to help fuel AI data center demand.
The benefit for enterprise IT is that as AI projects move from the planning to the implementation phase, hyperscalers will have the capacity to handle the huge data sets associated with large language model (LLM) workloads. On top of that, new public cloud data centers mean more availability zones, and fewer latency and performance issues associated with users not being physically close to the source of cloud-based applications and data.
Industry clouds are specialized cloud environments tailored to meet the unique requirements of specific sectors, offering pre-configured solutions.
In a recent Gartner survey, close to 39% of respondents said they had started the adoption of industry cloud platforms, with another 14% in pilots. Overall, a majority of respondents familiar with the concept identified themselves as adopters or potential considerers of industry-specific clouds.
A variety of players offer industry clouds. For example, AWS and Azure offer healthcare clouds, Salesforce has a financial services cloud, Oracle has a retail cloud, Siemens has a manufacturing cloud, and Google has an education cloud. Most hyperscalers offer multiple vertical clouds, with numerous niche players involved in the market as well.
When evaluating cloud costs, there are two aspects to consider: actual cost and anticipated cost. Many organizations who moved apps to the cloud because they thought it would be cheaper were shocked when the bills started arriving. In fact, cloud analyst David Linthicum estimates that cloud costs are coming in at 2.5 times what organizations anticipated.
On top of that, organizations are constantly expanding their cloud footprint.Mayank Bhargava, vice president and cloud modernization practice leader at CGI, a business and IT consulting firm, says, “As organizations continue their digital transformation journeys, they are migrating more critical workloads and data to the cloud, which naturally leads to higher costs.”
But there are things that organizations can do to get a handle on cloud costs. According to Forrester Research, the cloud cost management and optimization (CCMO) market is experiencing “meteoric growth.” These tools provide visibility into all cloud costs across hybrid and multicloud environments, identify opportunities to optimize cloud spend, and automate some actions.
CCMO can be viewed as a stepping stone to a more comprehensive and cross-departmental approach known as FinOps, which ties in finance and business units. FinOps teams work to optimize cloud costs, but also to perform budgeting, forecasting, and reporting that links cloud spend with business outcomes. Some vendors offering FinOps include Apptio, Lucidity, Densify, and Finout.
Enterprise IT is under pressure to reduce data center carbon emissions. Shifting resources to the cloud doesn’t get you off the hook; enterprises must track and report sustainability efforts across multicloud environments.
AI throws an unwelcome wrinkle in sustainability efforts because GPU chips run significantly hotter that traditional CPUs and therefore require more cooling, which spells more demand for electricity.
But hyperscalers are taking a leadership role, leveraging their advantages over enterprise IT. After all, data centersare their business, so they have the motivation, C-suite support, technical expertise, and financial resources. Unlike most businesses tethered to a specific physical location, hyperscalers can also scour the globe for sites that offer sustainably sourced energy.