Here's the thing about the AI boom: everyone's racing to build bigger, smarter systems, but nobody wants to melt their data centers in the process. Enter Arm Holdings plc (ARM), which is quietly positioning itself as the power-efficiency backbone of the entire artificial intelligence infrastructure buildout.
The Cambridge, U.K.-based chip designer has turned its Neoverse platform into what amounts to the hyperscaler's favorite energy-saving play. While everyone fixates on raw compute power, Arm is selling something more practical: "intelligence per watt." It's the kind of metric that matters when you're running warehouse-sized server farms and the electric bill arrives.
The strategy is working. Arm is targeting 50% market share among major hyperscalers in 2025, and the customer roster already reads like a who's who of cloud computing. Amazon.com, Inc (AMZN) Amazon Web Services is using Neoverse. So are Alphabet Inc (GOOGL) Google, Microsoft Corp (MSFT), Oracle Corp (ORCL), and Meta Platforms, Inc (META). Even next-generation AI campuses like the Stargate project are building their compute infrastructure around Arm for power efficiency and scalability.
The Nvidia Connection
To handle the surge in demand, Arm is expanding Neoverse with Nvidia Corp's (NVDA) NVLink Fusion—the high-performance interconnect that powers Grace Hopper and Grace Blackwell systems. This isn't just a minor technical upgrade. It's bringing Grace Blackwell-class performance to every partner building on Neoverse.
Arm CEO Rene Haas emphasized the collaboration's significance, while Nvidia CEO Jensen Huang called NVLink Fusion "the connective fabric of the AI era," describing how it unifies CPUs, GPUs, and accelerators at rack scale. The partnership builds on what Arm and Nvidia first achieved with Grace Hopper two years ago—coherent CPU-GPU integration. Now, extending NVLink Fusion across the broader Neoverse ecosystem gives chipmakers flexibility to pair Arm compute with their preferred accelerators, improving bandwidth and energy efficiency while cutting integration time.
Arm is also enabling Neoverse with its latest AMBA CHI C2C protocol—a technology Arm invented—to ensure seamless interoperability with NVLink Fusion. By eliminating memory bottlenecks and increasing data flow between components, partners can build differentiated AI systems that scale to meet the demands of this new compute era.
Financial Performance
The market is paying attention. Arm shares have gained 14% year-to-date. On November 5, the British chip designer reported second-quarter results that exceeded expectations, sending shares more than 4% higher in after-hours trading.
The company posted earnings of 39 cents per share, beating the 33-cent consensus estimate. Revenue climbed to $1.13 billion, ahead of forecasts and up from $844 million a year earlier. Growth showed up across key business lines. Royalty revenue increased 21% to $620 million, while licensing and other revenue surged 56% to $515 million. Arm also reported annualized contract value of $1.6 billion, a 28% jump year-over-year, reflecting sustained demand for its technology across cloud, AI, and edge markets.
ARM stock closed higher by 0.35% at $140.26 on Monday, continuing its steady climb as the company establishes itself as essential infrastructure for the AI buildout.