Data Center Demand Driven by AI Innovation

Lisa Su, Chair & CEO at AMD is highlighting a major shift happening in AI infrastructure right now: CPUs are becoming critical again as companies move from simple AI chatbots to full AI-driven workflows and agentic AI systems. Enterprises are rapidly increasing adoption of AI across software development, automation, analytics, and enterprise workflows — and that surge is driving unexpectedly high CPU demand alongside GPUs. A key reason is that modern AI workflows are no longer just “GPU problems.” AI agents now: orchestrate tasks, retrieve data, run simulations, compile code, manage inference pipelines, coordinate multiple models, and handle real-time enterprise operations. Those orchestration and infrastructure layers rely heavily on CPUs. AI server deployments are shifting from traditional CPU-to-GPU ratios like 1:8 toward configurations closer to 1:1. AMD’s data center business reflects that trend: Q1 2026 data center revenue grew 57% year-over-year to $5.8B. AMD forecasts server CPU revenue growth above 70% YoY in Q2. The company now expects the server CPU market to grow at a 35% CAGR through 2030, reaching around $120B. AMD is also positioning itself across the full AI stack: AMD EPYC CPUs for orchestration and inference, AMD Instinct GPUs for training, Ryzen AI for edge and enterprise AI PCs, plus networking and rack-scale AI systems. One of the most important insights from Lisa Su’s comments is that AI adoption is moving from experimentation to operational deployment. Companies are no longer testing AI in isolated pilots — they are embedding AI into real workflows, and that dramatically increases compute demand across CPUs, GPUs, memory, storage, and networking. For the tech industry, this signals a broader transition: AI infrastructure is evolving from “GPU-centric” to “full-stack compute architecture.” via @cnbctv #AI #ArtificialIntelligence #AMD #AIInfrastructure #DataCenter #EPYC #RyzenAI #MachineLearning #GenerativeAI #AgenticAI #EnterpriseAI #CloudComputing #Semiconductors #TechLeadership #DigitalTransformation #FutureOfWork #Innovation

Happy to finally share BloombergNEF US Data Center Outlook. This report combines our AI data center primer and US forecast into one incredible deep dive. We left no server rack unchecked – from AI training model demands to project construction timelines – this outlook covers it all. Key Findings: ⚡ BNEF projects US data-center power demand will more than double by 2035, rising from 34.7GW today to 78.2GW. Meanwhile, energy consumption could nearly triple, with average hourly electricity demand jumping from 16.2GWh to 49.1GWh. PJM is expected to remain the biggest by 2035 –followed by Ercot and then the Southeast. ⚡ BNEF’s relatively conservative forecast isn’t downplaying AI – it simply factors in real-world constraints like interconnection delays and build timelines. In the US, a data center takes seven years to reach full operation. For interconnections alone, developers face waits of 2–3 years in Chicago or 7–11 years in parts of Virginia and Texas. ⚡ Four companies – Amazon Web Services (AWS), Google, Meta and Microsoft – currently control 43% of US data-center capacity in 2024, wielding substantial influence over energy infrastructure planning and investment. ⚡ Data-center location decisions hinge many things like power cost, clean power, workforce availability, and tax incentives. But in the age of AI, speed-to-market and scalability top the list. Some developers co-locate near power plants or stranded renewables; others use remote campuses with bridging technologies to accommodate massive AI workloads. Read more here: https://lnkd.in/gAcgP9it Special thanks to Nathalie Limandibhratha (our lead author), along with Tom Rowlands-Rees, Jennifer W., Ben Vickers, and Ashish Sethia, for the many hours and dedication that made this note possible. And to our global counterparts – Jinghong Lyu, Ian Berryman, and David Hostert – it has been a pleasure to hack this data center topic together. What's in the report? ▪️ Section 1: Key findings on growth, AI’s role and hyperscaler influence. ▪️ Section 2: Basics of data-center types, components and efficiency metrics. ▪️ Section 3: How AI training drives massive power needs, cost and design shifts. ▪️ Section 4: Factors influencing where data centers are built. ▪️ Section 5: Regional analysis of major and emerging US data-center markets. ▪️ Section 6: BNEF’s demand and capacity outlook through 2035. Looking to dive deeper into the data?
 The downloadable Excel (included with this report) features: ▪️ All charts & underlying data from the study ▪️ US-wide, project-level data covering every operating data center (April 2025) ▪️ County-level data on pipeline capacity for data centers (April 2025)

The data center landscape is rapidly evolving with the emergence of "AI factories." These data centers are dedicated to a single application by a single customer, a shift from the traditional multi-tenant model. NVIDIA CEO, Jensen Huang, highlighted this trend, where these AI factories process data, train models, and generate AI for specific clients. The key drivers for AI factories are the power-intensive nature of AI processing, requiring specialized data centers with high-density and liquid cooling. These facilities might be strategically located in urban areas or repurposed existing spaces with ample power availability. While they're not widespread yet, the growing demand for AI suggests they may become more common. Security and regulatory concerns around AI could also lead to dedicated AI clusters that are required to comply with data sovereignty legislation. AI factories represent a new trend in data centers, catering to the unique demands of AI applications, and they are poised to grow alongside the AI industry's expansion. To expand on the AI data centre models a little more, we are supporting two distinct needs in the AI landscape. ‘Training’ data centers and ‘inference’ data centers, though both pivotal in the AI ecosystem, serve distinct functions. Firstly, AI training data centers are where the heavy lifting happens. They're akin to a gym where AI models, like athletes, train and develop. The process involves feeding vast amounts of data to the AI algorithms to teach them how to make predictions or take actions. This training phase requires substantial computational power and resources because the models need to process and learn from huge datasets. It's resource-intensive, both in terms of computational needs and energy consumption. On the other hand, AI inference data centers are more like the field where the trained athletes (AI models) play. Once the AI models are trained, they're deployed in these data centers to apply what they've learned to new data. This is called 'inference'. Here, the AI models make predictions or decisions based on the input data they receive. The computational load in inference data centers is typically lighter compared to training data centers. The focus is more on speed and efficiency, providing quick responses to the input data with minimal latency. In essence, training data centers are about building and teaching the AI models with heavy computational demands, while inference data centers are where these trained models are applied, focusing on speed and efficiency. Each plays a crucial role in the lifecycle of AI applications. Whatever the model - the size, scale, power density and cooling requirements are an order of magnitude larger than we have seen before and driving a huge change in the future of data centre design and operation. #whereailives #ai #aifactories

✍️ 𝗣𝗼𝘄𝗲𝗿𝗶𝗻𝗴 𝗔𝗜: 𝗔 $𝟮 𝗧𝗿𝗶𝗹𝗹𝗶𝗼𝗻 𝗖𝗵𝗮𝗹𝗹𝗲𝗻𝗴𝗲, 𝗮𝗻𝗱 𝘄𝗵𝗮𝘁 𝗶𝘁 𝗺𝗲𝗮𝗻𝘀 𝗳𝗼𝗿 𝘁𝗵𝗲 𝗨𝗦, 𝗠𝗘𝗡𝗔, & 𝗖𝗜𝗦. 🌎 🙋♂️ Through my ventures in data center investments and business development across the MENA, US, and CIS regions, including my work with AI-driven healthcare initiatives, I've seen firsthand the escalating demands for AI infrastructure. 📣𝙒𝙝𝙞𝙡𝙚 𝙗𝙞𝙡𝙡𝙞𝙤𝙣-𝙙𝙤𝙡𝙡𝙖𝙧 𝘼𝙄 𝙙𝙚𝙖𝙡𝙨 𝙢𝙖𝙠𝙚 𝙙𝙖𝙞𝙡𝙮 𝙝𝙚𝙖𝙙𝙡𝙞𝙣𝙚𝙨, 𝙖 𝙘𝙧𝙞𝙩𝙞𝙘𝙖𝙡 𝙘𝙝𝙖𝙡𝙡𝙚𝙣𝙜𝙚 𝙡𝙤𝙤𝙢𝙨: ‼️We're facing an $800 billion revenue shortfall for data centers, necessitating an estimated $2 trillion in investment by 2030 to maintain the current pace. It isn't just growth; it's a gold rush for computing power, the new most valuable commodity. 🪄 Consider these points: ✔️ AI's compute demand is doubling at twice the rate of Moore's Law, a pace of progress we've never seen before. ✔️ AI server growth is projected at a 41% CAGR, driving the overall data center market to a 23% CAGR. ✔️ To meet this demand, we need to invest $500 billion annually in data centers over the next decade. ✔️ The cost to construct a data center building has surged by 322% in just four years, before even adding a single chip or server. 📍 In healthcare AI—a sector I've focused extensively on in the MENA region—the infrastructure demands are particularly acute. Medical imaging, AI, and genomics processing require sustained high-performance computing, making reliable, cost-effective data center access critical for healthcare innovation. ♾️ This explosive growth is creating a significant energy bottleneck. Power demand from AI centers is set to quadruple in the next decade. By 2035, they could consume 1,600 terawatt-hours of power, equivalent to 4.4% of global electricity demand. 🔎 The AI revolution is still in its early stages. Addressing this $2 trillion challenge requires collaboration among investors, technology innovators, energy providers, and policymakers worldwide, from the US to the CIS and from Europe to the MENA region. 🖍️ 𝗔𝗱𝗱𝗿𝗲𝘀𝘀𝗶𝗻𝗴 𝘁𝗵𝗶𝘀 𝗰𝗵𝗮𝗹𝗹𝗲𝗻𝗴𝗲 𝗿𝗲𝗾𝘂𝗶𝗿𝗲𝘀 𝗮 𝗳𝗼𝘂𝗿-𝗽𝗿𝗼𝗻𝗴𝗲𝗱 𝗮𝗽𝗽𝗿𝗼𝗮𝗰𝗵: ↗️ Alternative energy partnerships (nuclear, renewable microgrids), ↗️ Next-generation cooling technologies (liquid cooling, immersion cooling), ↗️ AI-optimized chip architectures that improve performance per watt, and ↗️ Strategic geographic distribution to leverage regional energy advantages. ⁉️ The future of AI depends on the physical infrastructure we build today. Which emerging markets do you see as most promising for sustainable AI infrastructure development? 🧐 How are you balancing immediate scaling needs with long-term sustainability commitments? 📶 Let's connect and discuss the future of AI infrastructure. #DataCenters #AI #Investment #Energy #PhysicalAI #AICenters #MENA #CIS #Healthcare #Data #Infrastructure

🤖 Data Center Power Demand in the Age of AI The rapid expansion of artificial intelligence (AI) is driving a structural increase in global electricity demand. According to the IEA, electricity consumption from data centers is projected to nearly triple globally by 2035, rising from ~500 terawatt-hours (TWh) in 2025 to ~1,300 TWh. For context, that’s more than the current electricity use of entire countries like Japan or Germany. The primary drivers of this electricity consumption are training AI/ML models, cloud services, and digital infrastructure that require unprecedented computing power. Breaking it down, the U.S. and China dominate this expansion. By 2035, U.S. data centers alone are expected to consume ~600 TWh. China follows closely, with its data center demand rising steeply, reflecting its ambition to lead in AI adoption and digital services. While renewables are growing, coal, natural gas, and nuclear will remain significant in powering these massive digital engines. This creates both opportunities and risks — especially as governments and companies balance the need for energy security with environmental impact. 📣 Bottom Line The AI buildout is rapidly becoming an energy problem. Compute requires power, and power requires infrastructure. For investors, this shift signals opportunity far beyond semiconductors — spanning utilities, grid modernization, nuclear, natural gas, battery storage, and renewable developers. In the age of AI, electricity is the new strategic resource. ✅️ Key Insights 1) AI is driving a structural jump in electricity demand. Global data center power consumption is expected to surge from ~500 TWh in 2025 to over 1,300 TWh by 2035 — comparable to the electricity usage of major industrialized nations. 2) Renewables will carry much of the load. Clean energy sources are projected to supply a large share of incremental demand, highlighting how AI growth is increasingly tied to renewable deployment. 3) Natural gas remains the reliability backbone. Despite decarbonization goals, natural gas is likely to remain critical for grid stability due to its ability to deliver dispatchable power when renewables fluctuate. 4) China and the U.S. are entering an infrastructure race. Both countries are dramatically expanding generation capacity to support AI leadership, turning electricity into a strategic competitive advantage. 5) Power availability may become the new AI bottleneck. In the coming decade, access to energy — not chips — could determine where the next wave of hyperscale data centers gets built.

With Great AI Comes Great Power (Demand) Grid operators are increasingly concerned as electrification accelerates, particularly in industrial heating and transportation (EVs), placing immense pressure on the grid. AI and datacenters will magnify this challenge significantly. A ChatGPT query consumes roughly 10 times the energy of a Google search. If all Google queries were replaced with ChatGPT-like interactions, total annual energy consumption would reach around 9 TWh—about 4% of U.S. data center energy use. Goldman Sachs projects global data center power demand will surge 160% by 2030, rising from 1-2% to 3-4% of total global power consumption. India's Data Center Growth & AI Impact India’s data center capacity is estimated at 950 MW (2024) and is expected to grow to 1,700 MW by March 2025, though this may not fully account for AI-driven power demands. AI power consumption can be categorized into: Training – The computationally intensive process of developing AI models. Inference – Running queries through trained models to generate responses. Inference is likely the dominant AI power demand driver in India, though foundational AI model development could change this. Large models like ChatGPT require significant energy for inference. Alphabet chairman John Hennessy estimated an AI query costs 10 times more energy than a standard search. India’s AI Data Center Expansion Avendus Capital estimates AI-driven data center capacity in India could rise by 500 MW in four years. India’s data center market doubled from 540 MW in 2019 to 1,011 MW in 2023 and is expected to grow at a CAGR of 26% over the next three years, making it one of the fastest-growing globally. Power Demand & Efficiency Gains AI power demand can also be estimated by analyzing GPU sales. It needs to noted though AI compute power will not be evenly distributed worldwide. Europe’s high electricity prices (nearly double those of the U.S.) limit its AI compute capacity to just 4% of the global total. A crucial overlooked factor is efficiency. Data centers are becoming more power-efficient, with Power Usage Effectiveness (PUE) dropping from 2.7 in 2007 to 1.5 in 2021, with leading facilities reaching as low as 1.1. Nvidia’s Blackwell GPUs are reportedly 25 times more efficient than predecessors. Professor Jonathan Koomey’s research, Koomey’s Law, suggests compute energy efficiency doubles every 18 months, aligning with Moore’s Law. The Bigger Picture for India While energy efficiency gains will mitigate some impacts, power tariffs will be crucial to attract AI compute investment. India’s industrial power costs remain among the highest globally, and will be a big determinant of how much AI compute power the country attracts. The AI revolution is here, but its energy footprint must be managed. Balancing power demand, power pricing, infrastructure expansion, and efficiency will determine how India navigates this transformation.

The rapid expansion of AI is poised to transform industries across the globe, with companies expected to invest approximately $1 trillion in the next decade on data centers and their associated electrical infrastructure. However, a significant bottleneck threatens to slow this growth: the availability of reliable power to support the computational demands of AI systems. Today’s AI workloads require immense processing capacity, which is stretching the limits of existing power infrastructure. These demands make it increasingly challenging to secure sufficient electricity to maintain current data centers and, in many cases, prevent the construction of new facilities. AI models are more energy-intensive than the previous cloud computing applications that drove data center growth over the past two decades. At 2.9 watt-hours per ChatGPT request, AI queries are estimated to require 10x the electricity of traditional Google queries, which use about 0.3 watt-hours each; and emerging, computation-intensive capabilities such as image, audio, and video generation have no precedent. The stakes are high. After more than two decades of relatively flat energy demand in the United States—largely due to efficiency measures and offshoring of manufacturing—total energy consumption is projected to grow as much as 15-20% annually in the next decade. A significant portion of this increase is attributed to the expansion of AI-driven data centers. If current trends continue, data centers could consume up to 9% of the total U.S. electricity generation annually by 2030, more than doubling their share from just 4% today. The increasing scale and complexity of AI deployments are forcing companies to confront the harsh reality of existing infrastructure limits. Amazon Web Services recently invested $500M in Small Modular Reactors (SMR), whose technology is not yet commercially operable and isn't anticipated to come online until 2030-2035. Google signed a $100M+ power purchase agreement with an early stage SMR startup that won't have a viable unit until 2030. Microsoft convinced Constellation Energy to restart Three-Mile Island nuclear plant with a 20 year power purchase agreement. Addressing this power bottleneck requires not only technical innovation but also a deep understanding of both the electrical utility landscape and the operational needs of large-scale technology deployments. The solution will not be one size fits all. There will be a combination of many solutions required to solve the short-term immediate gap and long-term infrastructure needs. It will most likely require some combination of the following: intentional locating of data centers, improvements in data center processing efficiency, temporary fossil fuel power generation (natural gas), SMRs and “behind the meter” power purchase agreements.

Simplifying AI for Everyone #27 Why Data Centers Are Becoming the New Energy Companies Everyone talks about AI models. Few talk about the infrastructure reality behind them. AI does not scale on algorithms alone. It scales on power, cooling, location, and resilience. Today, data centers are no longer neutral IT assets. They are becoming energy-intensive industrial systems. Here is the uncomfortable truth: • AI demand is exploding exponentially • Grid capacity is growing linearly • Sustainability targets are tightening • Energy costs are becoming strategic, not operational This changes everything. Modern data centers now sit at the intersection of: • Energy generation • Grid optimization • Advanced cooling • AI workload orchestration The winning data center is not the one with the fastest servers. It is the one that understands energy as a core design principle. This is where AI plays a different role. AI is no longer just consuming power. It is becoming the brain that optimizes energy itself: • Predictive load balancing across regions • AI-driven cooling efficiency • Renewable energy forecasting • Smart workload shifting based on energy availability • Carbon-aware compute scheduling In other words: The future data center is an AI-managed energy system that happens to run compute. This shift is especially relevant in regions like the Middle East, where: • Renewable energy scale is unmatched • Land availability enables new architectures • Sovereign compute is a national priority • Energy cost competitiveness is strategic The next generation of data centers will not be built by IT companies alone. They will be built by energy + technology partnerships that understand both worlds deeply. And those who design for this reality today will own the infrastructure advantage tomorrow. #AI #DataCenters #Energy #DigitalInfrastructure #Sustainability #RenewableEnergy #AIInfrastructure #Vision2030 #Leadership #FutureOfCompute

The rapid global adoption of Artificial Intelligence is fueling a substantial increase in energy demand, largely driven by the need for expanding data centers to train and operate sophisticated AI models. While AI offers tremendous potential for innovation and societal progress, its burgeoning energy appetite poses a significant hurdle in the pursuit of a sustainable and decarbonized future. Let's have a look at AI's Energy Consumption by the Numbers: (1) AI is projected to trigger a 160% surge in data center power demand between 2022 and 2030, contributing considerably to the overall 2.4% increase in US power demand during that period. (2) By 2030, AI-driven data centers are expected to account for 8% of US power consumption, a significant jump from 3% in 2022. (3) The computational power required for AI is doubling roughly every 100 days, indicating an accelerating demand for energy. Training advanced models like GPT-4 already consumes an estimated 50 times more electricity than its predecessor, GPT-3. (4) Projections suggest that by 2028, AI's energy consumption could surpass that of entire countries, highlighting the scale of its impact on the energy landscape. The increasing energy demands of AI, coupled with population growth and electrification trends, are placing immense strain on the electrical grid. Ensuring grid stability and resilience in the face of this growing demand is a complex challenge requiring a multifaceted approach. While AI's energy consumption is a cause for concern, it also presents significant opportunities to optimize energy use and accelerate the transition to a clean energy future. AI's ability to analyze vast datasets and forecast energy production can be leveraged to enhance grid stability, optimize energy consumption, and integrate renewable energy sources more effectively. However, benefits should be probably be limited vs required energy demand. Addressing the intricate relationship between AI's energy use, its environmental impact, and its societal benefits requires a collaborative effort across industries and sectors. The World Economic Forum Artificial Intelligence Governance Alliance is actively working to establish a cross-industry framework to harness AI's potential for positive transformation while mitigating its energy footprint. As AI continues to evolve and permeate various aspects of our lives, finding sustainable solutions to its escalating energy demands is imperative. Embracing innovation, prioritizing energy efficiency, and strategically harnessing AI's capabilities can pave the way for a future where technological advancement and environmental sustainability go hand in hand.

The #AI Economy is fast growing — But Are Our #DataCenters Ready? There is enough literature online on the business impact of AI but are we forgetting about the need to reimagine our existing #DataCenters?? By 2030, AI is forecasted to be a $4 trillion #economy. But behind every intelligent #chatbot, #automation engine or #selflearning model lies an invisible force: #datacenters. And they’re under pressure like never before. #AI is #Power-Hungry—Literally Training advanced AI models demands massive #compute power. We’re talking high-performance #GPUs, parallel processing, and #petabytes of data. The consequence? A surge in #energy consumption and heat generation. Today, data centers globally consume 70GW of power. In just five years, that figure is projected to triple to 220GW To put in perspective: 1GW = 100 million LED bulbs. A staggering 75% of this growth will be driven by AI workloads alone. Rethinking #Location & #Architecture The AI boom is decentralizing data center geography #Tier1 markets are saturated. Enter #Tier2 and #Tier3 regions like #Utah, #Wyoming — areas with surplus energy, open space and lower local consumption But it’s not just about where we build. It’s about how we build. #Traditional #architectures won’t suffice High-density #AI #workloads demand #liquidcooling, efficient #airflow systems, and power-aware #chip design. In many centers, cooling alone eats up 30–40% of energy. The era of “just add more fans” is over. The #Power #Paradox The real constraint? #Energy. Not just delivering it—generating it. Power needs are escalating faster than our infrastructure can support. #Construction timelines stretch 18–24 months, often longer. Meanwhile, the supply chain is strained, and reliable sources of power are limited. #Innovation must bridge the gap. #Short-term: High-capacity battery storage, micro turbines #Medium-term: Small-scale nuclear reactors, gas plants #Long-term: Carbon capture, geothermal, solar mega-farms But all of this requires #capital — significant, upfront—and often carries uncertain #ROI. #Strategy Beyond Infrastructure This isn’t merely about tech adoption. This is about confronting structural limitations in a new AI-driven architecture. It demands we rethink the #fundamentals: Start from #consumption, not supply. Understand the demand profile: training vs inference, B2B vs B2C, demographics #Design for #flexibility: adaptable to changing supply chains, geopolitical dynamics, and macroeconomic shifts The #Ecosystem Imperative In this evolving landscape, partnerships will win. No single player— #hyperscaler or enterprise—can go it alone. Ecosystem collaboration across energy, infrastructure, and software is not optional—it’s #foundational. The #BigQuestion: As the AI economy surges ahead, can our data centers keep pace—or will they become the #bottleneck of tomorrow? What say you? #Sustainability #Innovation #Infrastructure #Cloud #EdgeComputing #FutureOfAI Data points credit McKinsey.