Arm CEO Says Data Center Business Is 'Exploding'

ARM Earnings Call Analysis - Q1 2024 & Future Outlook

Key Concepts:

• ARM Holdings: A semiconductor and software design company specializing in ARM architecture, widely used in mobile devices and increasingly in data centers.
• Hyperscalers: Large-scale data center operators like Google, Amazon, Microsoft, and Meta.
• Graviton & Grace (NVIDIA): ARM-based CPUs designed by NVIDIA for data center applications.
• Agentic AI: AI systems capable of autonomous action and decision-making, requiring significant CPU processing power.
• Stargate Initiative: ARM’s $500 billion investment plan focused on AI infrastructure.
• DPU (Data Processing Unit): A programmable processor that offloads tasks from the CPU, improving data center efficiency.
• Homogeneous Architecture: A data center infrastructure utilizing a consistent architecture (in this case, ARM) across all components.
• Free Float: The portion of a company’s shares available for public trading.

1. Financial Performance & Market Position

ARM reported record revenue of $1,240,000,000 and record royalties of nearly $740,000,000 for the quarter, representing a 27% year-on-year increase. The data center business experienced explosive growth, increasing by over 100% year-on-year and is projected to become ARM’s largest business within a few years. Despite concerns about the global memory shortage impacting the smartphone market, ARM has seen minimal direct impact due to its diversified business portfolio and the nature of royalty structures (memory shortages primarily affect the lower end of the market, which carries lower royalty rates). Guidance for the next quarter has been raised, reflecting continued strong performance.

2. Data Center Growth & the Handset-to-Data Center Transition

A key focus of the discussion was the accelerating shift from ARM’s traditional handset-centric business to a data center-focused model. ARM anticipates this inflection point – where data center revenue surpasses handset revenue – will occur “in a few years,” potentially sooner than previously expected (even six to nine months earlier than prior projections). This growth is driven by two primary factors:

• Increased ARM CPU Market Share with Hyperscalers: ARM now holds over 50% market share of CPUs within hyperscaler data centers.
• Higher Core Counts in ARM CPUs: CPU core counts are increasing significantly in ARM-based data center processors. Examples cited include Graviton chips moving from 96 to 192 cores and NVIDIA’s Grace CPUs increasing from 72 to 88 cores. This translates directly into higher royalty revenue for ARM.

3. The Role of AI in Driving Data Center Demand

The surge in data center demand is directly linked to the growth of Artificial Intelligence (AI), particularly “agentic AI.” The discussion highlighted that complex AI workloads, involving agent management, workflow orchestration, and general-purpose compute, are uniquely suited to CPUs. This is driving demand for more powerful ARM-based CPUs within data centers. Renee James stated, “When you think about what could be beyond AI, it's really hard to imagine,” emphasizing the long-term potential of the AI market. The $500 billion Stargate initiative, announced in January 2024, initially met with skepticism, but now appears increasingly credible given the scale of investment from major players like Google, Meta, and Microsoft.

4. NVIDIA’s Vera CPU & the Shift Towards Homogeneous Architectures

NVIDIA’s announcement of selling its Vera CPU as a standalone product is considered a significant positive development for ARM. The rationale is that data centers are moving towards “homogeneous” architectures, where a single architecture (ARM) is used across all components – CPUs, GPUs, storage, and DPUs – for ease of maintenance, upgrades, and cost efficiency. The Vera Rubin platform utilizes six times the number of CPUs compared to Grace Blackwell, highlighting the increasing importance of CPUs in modern data center designs. This shift is expected to further boost ARM’s royalty revenue.

5. SoftBank’s Stake & Stock Float

Renee James addressed concerns about SoftBank potentially selling its stake in ARM. She stated that SoftBank is a “very large shareholder who loves ARM” and has expressed no intention of selling. While the impact of SoftBank’s ownership on the stock’s “free float” was acknowledged, James indicated it is not a primary focus for the company. She emphasized ARM’s strong performance since its IPO in 2023, exceeding initial investor expectations.

6. Custom Chip Development & ARM’s Role

ARM is involved in conversations with numerous companies, including NVIDIA, OpenAI (potentially with Broadcom), and Meta, regarding custom chip development. A common theme is heavy investment in AI across various applications, from data centers to wearables and autonomous vehicles. ARM’s programmable, flexible, and low-power architecture positions it favorably to support these diverse workloads. James noted the challenge of predicting future AI architectures, but emphasized that ARM’s adaptability makes it well-suited to accommodate evolving needs.

7. Capital Expenditure & AI Infrastructure Buildout

The discussion addressed concerns about a potential “AI bubble” and bottlenecks in AI infrastructure. While acknowledging the significant capital expenditure (CapEx) increases from companies like Google (potentially $185 billion), James emphasized that this investment is driven by the immense opportunity presented by AI. She noted that the scale of investment now exceeds initial projections, suggesting a sustained commitment to AI development.

Notable Quotes:

• Renee James: “The common theme we see is continued investment in AI and not just at the data center. It's trying to figure out how do you run those AI workloads everywhere.”
• Renee James: “When you think about what could be beyond AI, it's really hard to imagine.”
• Renee James: “We are seeing such an investment because the AI opportunity, the way I like to think about it is kind of the final frontier relative to technology.”

Conclusion:

ARM is experiencing significant growth driven by the increasing adoption of its architecture in data centers, fueled by the demand for AI processing. The company is successfully transitioning from a handset-centric business to a data center-focused model, and its flexible, programmable architecture positions it well to capitalize on the evolving AI landscape. Despite concerns about the memory market and SoftBank’s ownership, ARM’s financial performance and future outlook remain exceptionally strong. The company is confident in its ability to continue exceeding investor expectations and solidifying its position as a key enabler of the AI revolution.