Microsoft is strategically reducing risk in its artificial intelligence (AI) expansion by securing contracts for graphics processing unit (GPU) capacity with its customers, a move revealed during its latest quarterly earnings call. This proactive approach signals a shift towards predictable AI infrastructure costs as demand surges.
Cloud Revenue Soars, AI Investments Continue
Table of Contents
Microsoft Cloud revenue exceeded $50 billion for the first time, growing 26% year over year.
- Microsoft Cloud revenue jumped 26% year over year, surpassing $50 billion.
- Overall revenue reached $81.3 billion, a 17% increase year over year.
- The company added nearly one gigawatt of datacenter capacity this quarter.
- Daily users of Microsoft 365 Copilot have increased tenfold.
- Microsoft anticipates a six-year useful life for its server and GPU fleet.
The company’s overall revenue climbed to $81.3 billion, up 17% year over year. Microsoft has significantly expanded its global datacenter capacity, adding almost one gigawatt this quarter. Alongside this expansion, the company reported a tenfold increase in daily users of Microsoft 365 Copilot, demonstrating growing adoption of its AI-powered productivity suite. Microsoft continues to invest heavily in AI infrastructure to meet the rising demand for Azure, Microsoft 365 Copilot, and GitHub Copilot.
Chief financial officer Amy Hood explained that “the majority of the capital that we’re spending today, and a lot of the GPUs that we’re buying, are already contracted for most of their useful life.” This strategy allows Microsoft to lock in pricing and availability, mitigating potential supply chain disruptions and cost fluctuations. She further clarified that GPU contracts are structured to cover the entire lifespan of the hardware.
Optimizing for Efficiency and Longevity
Microsoft’s AI infrastructure supports a diverse range of workloads. The company is focused on optimizing for “tokens per watt per dollar,” a metric that prioritizes hardware utilization and reduces total cost of ownership (TCO) through advancements in silicon, systems, and software. This pursuit of efficiency is crucial as AI models become increasingly complex and resource-intensive.
Microsoft anticipates a six-year useful life for its servers and GPUs. CEO Satya Nadella emphasized the role of software optimization in extending the lifespan of hardware. “We use software to continuously run even the latest models on the fleet,” he said. “That’s what gives us [longevity]. That’s why we think about ageing the fleet constantly. It’s not about buying gear one year – it’s about each year you ride Moore’s Law, you add and use software, and then you optimise across all of it.”
Following the introduction of Microsoft’s Maia 200 AI acceleration hardware, Nadella addressed the company’s focus on silicon development. He stated, “When you have a new workload, you can start innovating end to end between the model and the silicon and the entire system. It’s just not even about just the silicon.” He highlighted the importance of optimizing networking at the rack scale to accelerate memory access for demanding workloads.
Sovereignty and Shifting Customer Priorities
Alongside hardware and software investments, Microsoft is responding to a growing demand for data sovereignty. “Sovereignty is increasingly top of mind for customers, and we are expanding our solutions and global footprint to match,” Nadella noted. The company announced datacenter investments in seven countries this quarter, supporting local data residency requirements. Microsoft offers a comprehensive suite of sovereignty solutions across public, private, and national partner clouds, providing customers with tailored control over their data.
Senior analyst Dario Maisto observed, “The earnings call brought us back to the fundamentals after months of hype. In the cloud space, customers are now much more cautious about what they buy and how they pay for it.” He added that concerns about digital sovereignty are driving clients towards more selective and price-sensitive purchasing decisions, with increased consideration given to local cloud providers.
