• PE 150
  • Posts
  • Data Centers and AI

Data Centers and AI

How private equity is investing in AI Infrastructure, and the large Opportunity ahead

January 22, 2025 • Estimated Reading Time: 22 minutes

In this article

How private equity is investing in AI Infrastructure, and the large Opportunity ahead.

The AI revolution is advancing rapidly and requires critical infrastructure for continued growth.

Data centers form the backbone of this infrastructure, and private equity firms are strategically investing in this expanding market.

By doing so, they aim to capture future value and cash flows, leveraging the growing demand for data processing and storage capabilities driven by AI advancements, creating value within their portfolio.

How Large is the Generative AI Market?

According to Bloomberg Intelligence, Generative AI Global Market is expected to Reach a value of $1.3 Trillion by 2032E, from a 2022 base of $46 Billion, registering an impressive CAGR of 39.7% driven by training infrastructure in the near-term and gradually shifting to inference devices for large language models (LLMs), digital ads, specialized software and services in the medium to long term.

Bloomberg Intelligence estimates that Generative AI is poised to expand its impact from less than 1% of total IT hardware, software services, ad spending, and gaming market spending to 10% by 2032E. 

Largest drivers of incremental revenue: 

  • Generative AI infrastructure as a service ($247 Billion by 2032) used for training LLMs.

  • Digital Ads driven by the technology ($192 Billion)

  • Specialized generative AI assistant software ($89 Billion). 

On the hardware side: 

  • AI servers ($132 Billion).

  • AI storage ($93 Billion). 

  • Computer vision AI products ($61 Billion).

  • Conversational AI devices ($108 Billion).

Generative AI Market Opportunity Split

($ USD million)

2022

2032

2022 - 2032 CAGR

Hardware

$ 37,973

$ 641,737

33%

Devices

$ 4,128

$ 168,233

45%

  Computer Vision AI Products

$ 1,032

$ 60,564

50%

  Conversational AI Products

$ 3,096

$ 107,669

43%

Infrastructure (Training)

$ 33,845

$ 473,505

30%

  AI Server

$ 22,563

$ 133,817

19%

  AI Storage

$ 9,025

$ 92,642

26%

GenAI Infrastructure as a Services

$ 2,256

$ 247,046

60%

Software

$ 1,493

$ 279,899

69%

Specialized GenAI Assistants

$ 447

$ 89,035

70%

Coding, DevOps and GenAI   Workflows

$ 213

$ 50,430

73%

GenAI Workload Infrastructure Software

$ 439

$ 71,645

66%

GenAI Drug Discovery Software

$ 14

$ 28,343

114%

GenAI Based Cybersecurity Spending

$ 9

$ 13,946

108%

GenAI Education Spending

$ 370

$ 26,500

53%

GenAI Based Gaming Sports

$ 190

$ 69,414

80%

GenAI Driven Ad Spending

$ 57

$ 192,492

125%

GenAI Focused IT Services

$ 83

$ 85,971

100%

GenAI Based Business Services

$ 38

$ 34,138

97%

Total

$ 39,834

$ 1,303,651

42%

Source: Bloomberg

AI Market Share by Industry

According to Statista, the AI market share by industry presents a fascinating landscape of growth and investment. 

Leading the charge is the healthcare sector, which boasts an impressive 15.07% share, highlighting the burgeoning demand for AI-driven medical services and innovations. Following closely are the Finance and Manufacturing sectors, each capturing a 13.65% share, demonstrating the significant integration of AI in financial analysis, automation, and production processes. Business & Legal Services also show strong adoption at 13.60%, reflecting AI's role in enhancing business operations and legal processes. The Transportation sector commands a 10.75% share, underscoring the rise of AI in logistics and autonomous vehicles. Security applications, accounting for 9.90%, emphasize the critical importance of AI in safeguarding digital and physical assets. Other notable sectors include Media & Entertainment at 4.92%, Retail at 4.35%, and Energy at 5.29%, all of which are leveraging AI to optimize operations and enhance customer experiences. The Semiconductor industry, despite its foundational role in AI hardware, holds a smaller share at 2.39%, while the 'Others' category, encompassing various miscellaneous applications, stands at 5.81%. This diverse distribution showcases the widespread and varied impact of AI across different facets of the economy.

How Impactful can GenAI be to the World’s Economy?

Generative AI has the potential to significantly boost global productivity, potentially adding Trillions of dollars in value to the world economy. 

According to McKinsey, Generative AI could contribute between $2.6 Trillion and $4.4 Trillion annually across the 63 use cases the firm studied.

To put this into perspective, the entire GDP of the United Kingdom in 2021 was $3.1 Trillion. This increase would amplify the overall impact of artificial intelligence by 15% to 40%. 

Generative AI is poised to make a substantial impact across all industry sectors. Banking, High Tech, and Life Sciences are expected to experience the most significant effects as a percentage of their revenues. For instance, in the Banking sector, Generative AI could generate an additional $200 Billion to $340 Billion annually if fully implemented. 

Similarly, the Retail and Consumer packaged goods industries could see a substantial boost, with potential impacts ranging from $400 Billion to $660 Billion per year.

Generative AI has the potential to significantly enhance labor productivity across the economy.

However, realizing this potential will necessitate investments to support workers as they transition to new activities or jobs. 

Depending on the pace of technology adoption and the effective redeployment of worker time into other tasks, generative AI could drive annual labor productivity growth of 0.1% to 0.6% through 2040E.

Most Impacted Business Use Cases by AI Disruption

Generative AI has an outstanding potential to boost productivity and efficiency across all Business Sectors, helping organizations to drive decision-making faster and better.

Beyond its value in specific cases, generative AI can significantly benefit entire organizations by transforming internal knowledge management systems. With its advanced natural-language processing capabilities, generative AI enables employees to retrieve internal knowledge by posing queries as if they were conversing with a human. This continuous dialogue capability allows teams to quickly access pertinent information, facilitating faster and more informed decision-making and strategy development.

In 2012, the McKinsey Global Institute (MGI) estimated that knowledge workers spent about one day each work week, roughly 20% of their time, searching for and gathering information. Generative AI could dramatically enhance the efficiency and effectiveness of these workers by taking on these tasks. This virtual expertise can swiftly "read" extensive libraries of corporate information stored in natural language and scan source material in dialogue with a human, who fine-tunes and tailors the research. This approach is a more scalable solution than employing a team of human experts for the same purpose, offering substantial benefits.

Where will Value be concentrated?

As is typical with emerging technologies, investors are eager to discern which entities will thrive and falter, and where value will accumulate. In the realm of generative AI, a cluster of startups is emerging, each offering various foundational models. These startups have secured substantial funding from traditional VC sources and strategic investments from other technology firms, hinting at potential future mergers and acquisitions in this sector. 

Hyperscale cloud providers such as Amazon, Google, Microsoft, and Tencent, have forged partnerships with some of these foundational model providers, yet they also compete directly. Companies like Meta and ServiceNow have contributed their own open-source foundational models, offering competitive alternatives to closed-source models from companies like Anthropic, Cohere, and OpenAI.

Beyond these startups, the company that has seen the most substantial increase in value in recent times is GPU provider NVIDIA. At the beginning of 2023, NVIDIA's shares opened at $148.51 and surged to nearly +$6K today, with the company’s market cap stabilizing around $2.78 Trillion. Even before the launch of ChatGPT, there was a shortage of NVIDIA's GPUs, which are renowned for their efficiency in handling the parallel operations crucial for large-scale deep learning models.

Name

Total Raised

Most Recent Valuation

Selected Investors

AI21 Labs

$155M

$1.4B

Alphabet, NVIDIA, Samsung

Aleph Alpha

$142M

$0.49B

SAP

Anthropic

$7.66B

N/A

Amazon, Alphabet, Salesforce Ventures, Zoom Ventures

Cohere AI

$425M

$2.1B

Index Ventures, Tiger Global Management

Hugging Face

$395M

$4.5B

Salesforce Ventures, Alphabet, NVIDIA, Intel, AMD, QUALCOMM, IBM, Amazon

InflectionAI

$1.53B

$4B

Microsoft, NVIDIA, Greylock, Horizons Ventures

OpenAI

$11.3B

$86B

Microsoft, A16Z, Khosla Ventures, Sequoia Capital

StabilityAI

$151M

$4B

Sound Ventures, Lightspeed Ventures

Source: S&P Market Intelligence (2023).

AI Infrastructure: Datacenters

Datacenters serve as the foundational infrastructure for AI, offering essential components such as computational power, storage, scalability, connectivity, reliability, and specialized hardware. They play a vital role in facilitating the expansion and enhancement of AI applications across diverse industries.

Datacenters are pivotal to the evolution of AI for several crucial reasons:

Computational Power

AI algorithms, especially those involving deep learning, require significant computational resources. Datacenters provide the high-performance computing (HPC) infrastructure necessary to train and run these complex models efficiently. They house powerful processors, GPUs, and other specialized hardware optimized for AI workloads.

Storage

AI applications often deal with massive datasets, such as images, videos, and text corpora. Datacenters offer vast storage capacities to store and manage these large volumes of data. This storage capability is essential for training AI models, as well as for serving AI applications in real-time.

Scalability

AI projects often require scaling computational resources rapidly based on demand. Datacenters are designed to scale horizontally by adding more servers, GPUs, or storage units as needed. This scalability ensures that AI applications can handle increasing workloads without compromising performance.

Connectivity

Datacenters are equipped with high-speed networking infrastructure, enabling fast data transfer between servers and devices. This connectivity is crucial for AI applications that rely on real-time data processing or communication between distributed computing nodes.

Reliability and Redundancy

Datacenters are built with redundancy and reliability in mind. They employ backup power systems, cooling mechanisms, and security protocols to ensure uninterrupted operation. This reliability is essential for maintaining the uptime of AI services and preventing data loss.

Specialized AI Hardware

Some datacenters are deploying specialized AI hardware, such as TPUs (Tensor Processing Units) or AI-optimized ASICs (Application-Specific Integrated Circuits). These hardware accelerators are designed to perform AI computations more efficiently than traditional CPUs or GPUs, further enhancing the capabilities of AI systems.

Why data center industry operators need to keep up with growing AI demand

AI-powered applications, with their specific needs for computing power and storage, are poised to propel the next wave of growth in the data center industry. 

Operators are grappling with decisions regarding optimal locations, technical infrastructure needs, and future market dynamics. 

It is crucial for the data center industry to adopt proactive and adaptable strategies that encompass expansion flexibility, facility design, and energy management to navigate this evolving landscape effectively.

While AI has gained mainstream awareness, its size, reach, and impact are still in their infancy. This growth is putting pressure on the data center industry to anticipate and prepare for various scenarios related to performance expectations, government regulations, and the evolving shape of the AI industry.

The evolution of AI models

Source: EY.

Given the complexity demanded by Artificial Intelligence for reasoning and computations, the need for more and better data centers is crucial to support the rapid and significant advancements in this technology.

Topics like parallel supercomputing are among the clearest examples of the substantial demand for infrastructure generated by the revolution and evolution of Artificial Intelligence.

This demand creates a substantial market opportunity for data center managers, providers, builders, and investors supporting this capital-intensive sector of the economy, driving Private Equity Players to get involved in supporting this Long-Term Investments.

AI Impact in Data Centers Industry and Energy Demand

Goldman Sachs forecast a 15% CAGR in data center power demand from 2023 to 2030, driving data centers to make up 8% of total US power demand by 2030 from about ~ 3% currently.

Analysts now see a 2.4% CAGR in US power demand growth through 2030 from 2022 levels vs. ~0% over the last decade. Of the 2.4%, about 90 bps of that is tied to data centers.

Is estimated that 47 GW of incremental power generation capacity will be required.

to support US data center power demand growth cumulatively through 2030. This demand explosion is expected to drive ~$50 billion of capital investments in US power generation capacity in the mentioned period.

AI is expected to Boom Data Center`s Power Demand trough 2030

Case study: How Dominion is handling power demand growth from data centers

Dominion’s forecast of load from data centers indicates that the rapid growth it has seen will not slow down for the foreseeable future.

Why AI is so Impactful for Data centers?

AI is rapidly impacting the data center industry and power demand for several reasons:

  • Increased Compute Requirements: AI applications, particularly those involving machine learning and deep learning, require substantial computational power. This drives the need for more powerful and efficient data center infrastructure.

  • Data Storage and Processing: AI generates and processes vast amounts of data, increasing the demand for data storage and high-speed processing capabilities in data centers.

  • Extended Training Times: Training complex AI models can take days or even weeks, necessitating high-performance hardware and significant power consumption to handle the extended computational tasks.

  • Scalability Needs: AI-driven businesses often experience rapid growth, leading to a surge in demand for scalable data center resources to accommodate increased workloads and data volumes.

  • Energy-Intensive Operations: The hardware required for AI, including GPUs and specialized processors, is energy-intensive, leading to higher power consumption in data centers.

  • Increased Cooling Requirements: The rise in computer power also raises the need for advanced cooling solutions to manage heat generated by high-performance AI systems, further increasing energy demands.

ChatGPT queries are 6x-10x as power intensive as traditional Google searches

Power consumption per query/search (Wh).

US power demand Impact

Over the past decade, US power demand growth has averaged around 0% due to efficiencies such as LED lighting and the shift to more efficient data centers. However, power demand is expected to rise significantly in the coming years. The forecast for US power demand is now increased to 5,036 TWh by 2030, reflecting a 2.4% CAGR from 2022-2030, up from the previous estimate of 4,733 TWh and a 1.7% CAGR. This adjustment accounts for anticipated growth in data centers and AI. Residential, commercial (excluding data centers), and industrial demand are forecasted based on historical macroeconomic trends, while electric vehicle demand is projected separately.

Data center/AI growth adds about 80 bps on average to our annual power demand growth rates from 2024-2030

Increased capital investment

Goldman Sachs expects 2024-2027 capex to increase ~36% versus than the previous four-year period, and to reach a cumulative value of $563 Billion.

To support the anticipated growth driven by data centers and AI, significant investment in grid reliability, particularly in transmission, is required. The forecast includes a $720 Billion industry spending estimate through 2030, with approximately $260 Billion allocated to transmission and $465 Billion to distribution. This represents a 36% increase in capital expenditure from 2024 to 2027 compared to the previous four years. This surge in investment is expected to contribute to about 7% earnings growth across the covered utilities. Annual grid spending is projected at around $66 Billion for distribution and $37 Billion for transmission, reflecting a 38% increase in total capital investment over the next four years.

McKinsey estimates that Global spending on the construction of data centers will reach a value of $49 Billion by 2030, growing at a CAGR of 5.4%

Where is being Data Centers’ Infrastructure concentrated?

Northern Virginia leads the US in data centers, outpacing the next five markets combined. Growth is driven by low energy costs and tax incentives. 

Northern Virginia is the largest data center market in the US by a large margin.

Why is Northern Virginia the world’s data center capital?

  • Proximity to Major Markets

  • Robust Infrastructure

  • Low Energy Cost

  • High Capacity and Scalability

  • Reliable Power Supply

What are Private Equity Firms doing to capture this value?

Private equity's engagement with AI is more than just preliminary interest. Firms globally are eager to explore how AI can enhance their portfolio companies. Major players are investing in AI tools and teams to gain unique insights from proprietary data and potentially influence investment decisions in the future.

Simultaneously, private equity is heavily investing in data centers and related technologies, essential for supporting the AI boom. The ongoing shift to cloud computing and off-premise data storage has already driven strong demand for data centers, and the rapid advancement of AI is further intensifying this trend.

Source: Pitchbook.

Private equity is investing in data centers for AI infrastructure through several key strategies:

  • Direct Investments in Data Centers: Private equity firms are acquiring or building data centers to support the increasing demand for AI infrastructure, including both new developments and upgrades to existing facilities.

  • Funding Data Center Operators: Investments are being made in companies that operate data centers, which provide the essential infrastructure needed for AI and cloud computing.

  • Partnerships and Joint Ventures: Private equity is engaging in partnerships and joint ventures with data center providers and tech companies to create or enhance facilities optimized for AI workloads.

  • Technology Upgrades: Investment focuses on upgrading data center technologies to meet the high computational and storage demands of AI, including improved cooling and power management systems.

  • Investing in AI-Specific Infrastructure: Funding is directed towards infrastructure specifically designed for AI, such as edge computing facilities that process data close to its source to reduce latency.

  • Supporting Cloud Computing Expansion: Investments are made in cloud computing services and platforms that support AI workloads, enhancing scalability and efficiency for AI deployments.

North America and Europe: Leading the Way in Data Center Investments

In 2023, North America dominated data center investments, capturing a substantial 62% of global transaction values, and increasing to 69% of investments by April 2024, amounting to $15 Billion, with the United States leading the share.

Europe, on the other hand, has experienced a significant rise, boosting its market share from 6% in 2022 to 20% in 2023. By early 2024, Europe has attracted over $7 Billion in data center investments, holding a 29% share. This trend underscores Europe’s growing prominence in the data center sector.

While North America remains a major player, Europe is notable for achieving a year-over-year increase in transaction volume in 2023.

European Breakdown: France and Italy Rise as New Hotspots

France is becoming a notable player in the data center sector, with nearly $4 Billion invested through six M&A transactions in 2023. A significant portion of this investment came from Brookfield's purchase of the Data4 data center from AXA Investment Managers Real Assets for $3.8 Billion.

Italy is also emerging as a key investment destination, with over $1 Billion in greenfield investments across three projects. This includes the development of two 40 MW hyperscale data centers in Bellini and near Milan.

Over the past five years, the UK has attracted 30% of Europe's data center investments, totaling approximately $7 Billion.

What Experts Says on Data Centers AI-driven Trend

Investors and consultants are significantly increasing their allocations to data centers, driven by the growth of AI. 

Steven Meier, who oversees $278 billion in New York City retirement systems, is targeting investments in data centers, sustainable energy, and AI infrastructure. He sees substantial opportunities in debt strategies for these sectors and anticipates growth in venture capital and buyout investments.

The global infrastructure market is projected to grow by 27.5% from 2024 to 2033, reflecting the rising demand for data centers and AI technology.

Tim McCusker, CIO of NEPC, points out that AI is highly scalable, with significant growth occurring in China, Europe, and North America, and the Asia Pacific region expected to see the fastest expansion in AI over the next decade. 

The Blackstone Case

Blackstone, under CEO Stephen Schwarzman, aims to become the world's largest investor in AI infrastructure. This move is driven by a projected $2 Trillion in global capital expenditures for data center development. 

The firm sees significant opportunities in this sector due to the increasing demand for data processing and storage capabilities fueled by advancements in AI technology. This strategy aligns with Blackstone's broader investment approach, emphasizing large-scale, high-potential ventures.

Conclusion

AI continues its unstoppable growth, revolutionizing industries across the board. However, its expansion is constrained by two key factors: power and infrastructure. Private Equity sponsors are uniquely positioned to play a pivotal role in this evolving economy by deploying strategic growth capital to fund the development of AI infrastructure—capital-intensive investments that promise substantial mid- to long-term returns while enhancing portfolio value.

Beyond delivering attractive returns, Private Equity players will emerge as strategic enablers and key drivers of the next productivity frontier, solidifying their relevance in shaping the future economy.

Sources & References

Bloomberg. (2023). Generative AI to Become a $1.3 Trillion Market by 2032, Research Finds. https://www.bloomberg.com/company/press/generative-ai-to-become-a-1-3-trillion-market-by-2032-research-finds/ 

IMF. (2024). Global GDP.

Statista. (2024). Artificial Intelligence – Worldwide. https://www.statista.com/outlook/tmo/artificial-intelligence/worldwide