As detailed in the video below, we recently visited Japan, South Korea, Singapore and Indonesia to keep pace with the tremendous growth of digital infrastructure in the region.
Established operators such as Equinix, NTT and Singtel have found themselves particularly well-positioned to capitalize on this growth opportunity through their highly developed network infrastructure connecting major global cities. rice field.
Our own analysis below highlights one example of how Equinix, the host of Internet exchanges, network carriers and undersea cables that aggregate critical data, has a leading position in the Singapore market. The gravity of these major network nodes is evidenced by the number of networks peering to Equinix’s site. Simply put, the company makes high profit margins by interconnecting these tenants, and currently in Singapore Metro alone he hosts over 24,000 interconnections between network peers.
Data Centers: Modern Excavators in the AI Gold Rush
The mass adoption of generative artificial intelligence (AI), characterized by the overwhelming success of ChatGPT since November 2022, has sparked interest akin to the California Gold Rush. The investment market has jumped back on the Silicon Valley bandwagon, rewarding trailblazers such as NVidia, Google and Microsoft for gaining a first-mover advantage through extensive R&D efforts. Not surprisingly, investors are now looking for their next group of beneficiaries, trying to avoid speculative or even loss-making companies.
We believe the answer is hidden in plain sight. NVidia’s Q1 2024 earnings release marks a tipping point for AI supply chain, highlighting recent incredible growth in demand for NVidia’s hardware, analysts upping full-year earnings forecasts by nearly 40% Fixed (Source: Refinitiv).
When describing its upgraded outlook, the company mentioned “data centers” more than 56 times during its May 24, 2023 earnings call. Advanced graphics processing units (GPUs) accelerate the processing of large amounts of data, but the data required to train and apply AI models is completely dependent on a high-performance, secure and stable data center environment.
Evolution of data centers from communication exchanges to AI launchpads
Data centers have evolved dramatically from their origins as communication hubs to host the Internet in the late 90’s and early 2000’s. Technological and policy advances have since driven the transformation to larger, more localized and more resilient facilities with state-of-the-art cooling systems, redundant power supplies and advanced security measures. .
Over the past two decades, in developed markets (outside of Asia), these networking hubs have been spun off from telecom companies and reborn as long-term infrastructure businesses. These portfolios not only host the public internet, but also enable businesses to privately connect to their end customers and partners. Combined with the unprecedented surge in data generation from the smartphone revolution, the demand for data storage and processing power continued to skyrocket.
AI data science has been around since the 1950s, but it’s only recently that advanced AI technology has become affordable and widely available that adoption has skyrocketed. This is in no small part due to the fact that cloud computing giants such as Amazon and Microsoft are beginning to develop their own in-house applications and make them available to businesses of all sizes.
Quantify growth potential
While the size of the potential opportunity remains the subject of intense debate, total data center demand, defined by power consumption, could reach 35GW by 2030 from 17GW in 2022 in the US market alone. Prediction experts estimate that Global market share is about 40%. Analysts at Evercore ISI believe that future electricity consumption related to AI deployments could grow from about 1GW in 2023 to 7GW by 2026, representing a $12 billion revenue opportunity that could overwhelm existing We estimate a 15-20% increase compared to total data center capacity.
Equinix has a portfolio of 248 multi-tenant data centers in 32 countries, hosting more than 10,000 companies with over 450,000 interconnections between them. Beyond providing physical space, power, cooling, and connectivity, the company’s services increasingly focus on network services, giving it a competitive advantage.
At Analyst Day 2023, Equinix identified a Service Available Market (SAM) target of $21 billion in AI revenue by 2026 based on current operations and capabilities. They will probably only get a fraction of this SAM, but the potential for increased demand compares favorably with their current revenue run rate of about $8 billion annually (Source: Company, June 2023). ).
Can our power grid sustainably meet the demands of AI computing?
Even before the mass adoption of AI tools, metropolitan areas were struggling to keep up with the surge in demand for data center space. Due to supply chain bottlenecks and the scarcity of distributable power, data center owners in these locations are experiencing low vacancy rates and have a significant impact on pricing.
AI software is trained on vast amounts of data to return useful responses. These advanced applications use GPUs instead of central processing units (CPUs) because they can process multiple computations simultaneously and more efficiently. GPUs are more efficient per byte of data processed, but as suggested by the projections above, the total power consumed can increase as new use cases are introduced.
As data centers start hosting more high-power GPUs to support AI, the strain on the power grid will become even more acute. Some utilities have already set quotas that limit the construction of new facilities, increasing the value of existing ones. Singapore is a case in point, where power allocations for new data centers are tightly controlled by the government and disproportionately given to incumbents such as Singtel who can help arrange new solutions to their renewable power requirements.
AI supply chains must work hard to extract more efficiencies from their existing resources. R&D innovations in liquid cooling and optical networking technology are advances set to reduce the energy consumption required to run high-performance chips. Additionally, data center operators and occupiers are keen to add new, large-scale renewable energy supplies to the grid.
AI drives unprecedented demand for data centers
As the critical infrastructure behind the digital economy, data centers will play a key role in delivering new AI tools to consumers and businesses.
As with any new technology, the ultimate size of the AI market opportunity is still up for debate. Data center operators must carefully manage environmental and regulatory concerns. We believe that companies with a long track record of innovation should see even more revenue from this new wave of applications.
After an exhaustive survey of Asia’s major cities, we believe the region has the most exciting growth opportunities. Additionally, the “first movers” with the most established digital infrastructure networks are well positioned to capture the upcoming explosive demand.
source:
McKinsey & Company (2023). Investing in the emerging data center economy. McKinsey & Company.
Evercore ISI (2023), Equity Research Note
Nvidia Q1 Results May 2023
important information
Marketing Materials for Professional Clients Only
This document contains suggestive terms for descriptive purposes only and is not intended to provide the sole basis for evaluating the investment solutions described. Management of the portfolio may have been delegated (including partially delegated) to another Schroeder entity. Any reference to sectors/countries/stocks/securities is for illustrative purposes only and is not a recommendation to buy or sell any financial instrument/securities or to adopt any investment strategy. This material is not intended to provide accounting, legal or tax advice or investment recommendations and should not be relied upon.
Past performance is not a guide to future performance and should not be repeated. The value of your investment and the return you receive from it may go up or down, and you may never get your original investment back. Changes in exchange rates can cause investment values to rise as well as fall.
Schroders expresses its own views and opinions in this document, which are subject to change. Although the information contained herein is believed to be reliable, Schroders does not guarantee its completeness or accuracy.
Included projections are not guaranteed. They are provided only as of the date of publication and should not be relied upon. Our forecasts are based on our own assumptions and are subject to change. We accept no liability for errors of fact or opinion. We are also under no obligation to provide you with any changes to our assumptions or forecasts. Forecasts and assumptions may be affected by external economic or other factors.
The hypothetical results presented here should only be considered approximate representations of portfolio performance and are not indicative of historical performance. This is the result of statistical modeling based on many assumptions and has many serious limitations on retrospective reconstruction of performance results from performance records. For example, it does not take into account transaction costs or liquidity issues that may affect the performance of your actual investment. This data is provided for informational purposes only as of today and should not be relied upon to predict future performance. There is no guarantee that this performance could actually be achieved using the tools and data available at the time. No representation is made that any particular combination of investments was selected on the commencement date, held for the period indicated, or the performance achieved.
Third party data, including MSCI data, is owned or licensed by the data provider and may not be reproduced or extracted or used for any other purpose without the consent of the data provider. Third party data is provided without warranty of any kind. Data Providers and Document Publishers disclaim any liability with respect to Third Party Data. If there are specific third-party disclaimers, those terms are listed in the “Important Information” section of www.schroders.com.
Schroeder is the data controller with respect to your personal data. For information on how Schroders processes your personal data, please see our privacy policy available at www.schroders.com/en/privacy-policy/ or request if you cannot access this webpage. Please see accordingly. For security purposes, communications may be recorded or monitored.
Issued in the United Kingdom by Schroder Investment Management Limited (1 London Wall Place, London EC2Y 5AU). Registration number 4191730 UK. Licensed and regulated by the Financial Conduct Authority.
