Blab AI
AI-Driven Data Center Boom Drives Interest in Arm, IREN, and Nvidia
AI workloads are accelerating the construction of new data centers worldwide. Morgan Stanley Research estimates that global data‑center capital expenditure will reach about $2.9 trillion between 2025 and 2028, largely to support the next generation of artificial‑intelligence (AI) models. The forecast underscores a sustained need for compute capacity that current supply cannot meet.
Three companies are positioned to benefit from this trend. Arm Holdings, a British chip‑design firm, is expected to capture the largest share of data‑center central‑processing‑unit (CPU) sales by 2030. IREN Ltd., an Australian renewable‑energy‑powered data‑center operator, has secured long‑term contracts with Microsoft and Nvidia and is expanding its capacity to 5 GW of grid‑connected power. Nvidia Corporation, a U.S. GPU and system‑on‑chip (SoC) manufacturer, has seen its data‑center revenue double in the last quarter and is launching a new CPU line that could generate $20 billion in revenue this year.
The AI build‑out is already visible in the capital markets. According to Morgan Stanley, nearly $3 trillion of AI‑related infrastructure investment will flow through the global economy by 2028, with a large portion financed through debt. The estimate reflects the scale of the compute gap that AI developers and cloud providers face.
Arm’s architecture powers almost every modern smartphone and is increasingly adopted in servers. In the most recent quarter, Arm reported revenue of $1.49 billion, up 20 % year over year. The company earns royalties on every chip that licenses its architecture, a model that has proven highly profitable. Management forecasts that Arm will hold the largest share of data‑center CPUs by the end of the decade, citing its energy‑efficient designs and growing core counts. The firm’s market capitalization is about $380 billion, a figure that analysts say could rise substantially as data‑center demand for Arm‑based chips grows.
IREN’s growth is driven by its renewable‑energy‑centric data‑center model. The company’s stock has risen 385 % over the past year, while its market cap remains around $18 billion. IREN has signed two five‑year cloud contracts with Microsoft and Nvidia, valued at roughly $9.7 billion each. The contracts will provide Microsoft with access to Nvidia GPUs and IREN’s own GPU‑cloud services. IREN plans to bring 480 megawatts of new capacity online this year and expects $4.4 billion in annualized revenue by the end of 2026. Expansion plans include facilities in Spain and Australia, and the company aims to have 1.2 GW of its power portfolio online by 2027.
Nvidia’s data‑center business continues to expand. The company’s revenue from that segment nearly doubled in the last quarter, and its upcoming Vera CPU line is projected to reach $20 billion in revenue this year, representing more than a third of Intel’s revenue. Nvidia’s Vera Rubin platform, which combines multiple chip types to power agentic AI workloads, is expected to drive further growth. Analysts project that Nvidia’s total revenue could rise 81 % this year to $391 billion.
In summary, the AI‑driven demand for compute is creating a large capital‑intensive build‑out that is reshaping the data‑center landscape. Arm’s energy‑efficient CPU designs, IREN’s renewable‑energy‑powered facilities, and Nvidia’s expanding GPU and CPU portfolio are all positioned to benefit from the projected $3 trillion spend on data‑center infrastructure. The next few years will see continued investment in these companies as they scale to meet the compute needs of AI workloads.
Three companies are positioned to benefit from this trend. Arm Holdings, a British chip‑design firm, is expected to capture the largest share of data‑center central‑processing‑unit (CPU) sales by 2030. IREN Ltd., an Australian renewable‑energy‑powered data‑center operator, has secured long‑term contracts with Microsoft and Nvidia and is expanding its capacity to 5 GW of grid‑connected power. Nvidia Corporation, a U.S. GPU and system‑on‑chip (SoC) manufacturer, has seen its data‑center revenue double in the last quarter and is launching a new CPU line that could generate $20 billion in revenue this year.
The AI build‑out is already visible in the capital markets. According to Morgan Stanley, nearly $3 trillion of AI‑related infrastructure investment will flow through the global economy by 2028, with a large portion financed through debt. The estimate reflects the scale of the compute gap that AI developers and cloud providers face.
Arm’s architecture powers almost every modern smartphone and is increasingly adopted in servers. In the most recent quarter, Arm reported revenue of $1.49 billion, up 20 % year over year. The company earns royalties on every chip that licenses its architecture, a model that has proven highly profitable. Management forecasts that Arm will hold the largest share of data‑center CPUs by the end of the decade, citing its energy‑efficient designs and growing core counts. The firm’s market capitalization is about $380 billion, a figure that analysts say could rise substantially as data‑center demand for Arm‑based chips grows.
IREN’s growth is driven by its renewable‑energy‑centric data‑center model. The company’s stock has risen 385 % over the past year, while its market cap remains around $18 billion. IREN has signed two five‑year cloud contracts with Microsoft and Nvidia, valued at roughly $9.7 billion each. The contracts will provide Microsoft with access to Nvidia GPUs and IREN’s own GPU‑cloud services. IREN plans to bring 480 megawatts of new capacity online this year and expects $4.4 billion in annualized revenue by the end of 2026. Expansion plans include facilities in Spain and Australia, and the company aims to have 1.2 GW of its power portfolio online by 2027.
Nvidia’s data‑center business continues to expand. The company’s revenue from that segment nearly doubled in the last quarter, and its upcoming Vera CPU line is projected to reach $20 billion in revenue this year, representing more than a third of Intel’s revenue. Nvidia’s Vera Rubin platform, which combines multiple chip types to power agentic AI workloads, is expected to drive further growth. Analysts project that Nvidia’s total revenue could rise 81 % this year to $391 billion.
In summary, the AI‑driven demand for compute is creating a large capital‑intensive build‑out that is reshaping the data‑center landscape. Arm’s energy‑efficient CPU designs, IREN’s renewable‑energy‑powered facilities, and Nvidia’s expanding GPU and CPU portfolio are all positioned to benefit from the projected $3 trillion spend on data‑center infrastructure. The next few years will see continued investment in these companies as they scale to meet the compute needs of AI workloads.
