--- title: "Arm Enters the Fray Against Intel and AMD, First CPU Targets AI Data Centers, Projected Annual Revenue of $15 Billion" type: "News" locale: "zh-HK" url: "https://longbridge.com/zh-HK/news/280366078.md" description: "Arm's AGI CPU features up to 136 cores and targets x86 architecture CPUs, working in conjunction with AI accelerators. Initial customers include Meta. Arm stated that it is launching its own chip to \"capture a larger share of AI infrastructure spending.\" The company projects that the new chip business will generate over $15 billion in annual revenue within five years, surpassing its existing business, and bringing total annual sales to $25 billion, five times the current level. The stock price briefly turned higher, up 2.7%, before reversing down" datetime: "2026-03-24T20:14:27.000Z" locales: - [zh-CN](https://longbridge.com/zh-CN/news/280366078.md) - [en](https://longbridge.com/en/news/280366078.md) - [zh-HK](https://longbridge.com/zh-HK/news/280366078.md) --- > 支持的語言: [简体中文](https://longbridge.com/zh-CN/news/280366078.md) | [English](https://longbridge.com/en/news/280366078.md) # Arm Enters the Fray Against Intel and AMD, First CPU Targets AI Data Centers, Projected Annual Revenue of $15 Billion Chip design giant Arm is entering the arena as a competitor, officially releasing its first self-designed CPU, signifying a fundamental shift in its business model: moving from licensing fees to directly participating in the trillion-dollar AI computing market. On Tuesday, February 24th, US Eastern Time, Arm announced the launch of its first self-designed processor, named "AGI CPU," and confirmed Meta as one of its initial customers. The product, manufactured by TSMC, is aimed at AI data centers, featuring up to 136 cores and consuming approximately 300 watts, designed to enhance computing density and energy efficiency. Arm projects that the new chip business will generate approximately $15 billion in annual revenue within five years, surpassing the sales of its existing business and bringing the company's total annual sales to about $25 billion at that time, five times the current level. By then, Arm's earnings per share (EPS) are expected to reach $9. Analysts had previously forecast Arm's adjusted EPS, excluding certain items, for the current fiscal year to be only $1.75. This news quickly garnered market attention. Arm's stock price briefly turned higher in early and mid-morning trading on Friday, reaching an intraday high of about 2.7% in the afternoon before reversing down, with the decline once widening to over 2.8%, indicating clear divergence among investors regarding the prospects of its strategic transformation. By the close, Arm was down about 1.4%, then turned higher in after-hours trading. Analysis suggests that the rationale for the stock price increase is Arm's entry into the high-growth AI chip market, expanding its long-term revenue potential, and strengthening its position in data centers. The decline reflects uncertainties brought by the business model transformation, potential impact on existing customer relationships, and fierce competition from Intel, AMD, Nvidia, and others, making it difficult for the self-designed chip business to contribute significant profits in the short term. In the short term, Arm's self-designed chip is a high-investment, long-term strategic bet; in the long term, it may determine whether Arm can transition from a "behind-the-scenes winner" to a "front-stage giant." ## From "Selling Designs" to "Selling Chips": Arm's Key Business Model Leap For a long time, Arm's core business model has been to license its architecture and designs to chip manufacturers, achieving high gross profit growth through licensing fees and royalties. However, with the advent of the AI era, the value chain for data center chips has significantly shifted upwards, with the price per chip soaring from tens of dollars in the smartphone era to tens of thousands of dollars. Against this backdrop, Arm has chosen to enter the arena directly. Arm stated on Tuesday this week that the launch of its self-designed chip is aimed at "capturing a larger share of AI infrastructure spending." CEO Rene Haas directly stated that customer demand was a significant reason for the company's decision. The core logic of this transformation lies in: - The licensing model has extremely high profit margins (gross margins close to 98%), but limited revenue scale. - Data center chips, despite lower gross margins, operate in a massive market. - Arm aims to cut into higher-value segments while maintaining its licensing business. In other words, Arm is upgrading from a "technology provider" to a "product participant." ## AGI CPU: 136 Cores, Targets x86, Synergizes with AI Accelerators In terms of product positioning, AGI CPU does not directly target GPUs but serves as a "central hub" in AI infrastructure. According to the company's disclosures, the chip is primarily used for: - Coordinating computational tasks for AI accelerators (such as GPUs). - Handling data preparation and scheduling. - Running certain AI inference tasks. This means its positioning is akin to the "CPU foundation for data centers in the AI era," forming a synergy with accelerators. In terms of performance: AGI CPU features up to 136 cores, consumes approximately 300W, and focuses on higher energy efficiency. Arm emphasizes that its product outperforms traditional x86 architecture CPUs in energy efficiency, a key selling point for data centers where power and cooling costs are extremely high. ## Directly Confronting Intel and AMD, Challenging the x86 Camp Arm's move most directly impacts Intel and AMD, who have long dominated the server CPU market. The data center CPU market is currently still dominated by the x86 architecture, but Arm's architecture has seen rapid penetration in the cloud computing sector in recent years (e.g., AWS Graviton). Arm's introduction of its own chip means an escalation of competition: from "licensing Arm architecture vs. x86" to "Arm's self-designed chips vs. x86 vendors' products." Management has explicitly stated that its goal is to seize market share from the x86 camp. However, Arm is entering a highly crowded battlefield with equally fierce competition: Nvidia has already introduced its own CPUs, numerous AI chip companies like Cerebras are accelerating their deployments, and large cloud vendors like Meta are also developing their own chips. ## More Subtle Customer Relationships: Both Partner and Competitor Beyond technical challenges, a more complex issue lies in ecosystem relationships. Arm's traditional customers—including cloud vendors and chip design companies—almost universally develop products based on its architecture. Now, Arm is beginning to sell chips directly to these customers: - Meta is both an initial customer and a self-designed chip vendor. - Most major technology companies simultaneously procure solutions from Nvidia, AMD, and their own internal developments. - Arm may compete with its customers in certain scenarios. This model of "supplying technology while also selling products" could reshape industry partnerships. However, based on the current situation, major clients tend to adopt a "multi-vendor strategy": utilizing Arm, x86, and self-developed chips simultaneously to diversify risks and optimize costs. ### 相關股票 - [Arm (ARM.US)](https://longbridge.com/zh-HK/quote/ARM.US.md) - [Meta Platforms (META.US)](https://longbridge.com/zh-HK/quote/META.US.md) - [Intel (INTC.US)](https://longbridge.com/zh-HK/quote/INTC.US.md) - [Invesco AI and Next Gen Software ETF (IGPT.US)](https://longbridge.com/zh-HK/quote/IGPT.US.md) - [Direxion Semicon Bull 3X (SOXL.US)](https://longbridge.com/zh-HK/quote/SOXL.US.md) - [Invesco PHLX Semiconductor ETF (SOXQ.US)](https://longbridge.com/zh-HK/quote/SOXQ.US.md) - [Roundhill ARM WeeklyPay ETF (ARMW.US)](https://longbridge.com/zh-HK/quote/ARMW.US.md) - [Global X Data Center & Dgtl Infrs ETF (DTCR.US)](https://longbridge.com/zh-HK/quote/DTCR.US.md) - [GraniteShares 2x Long AMD Daily ETF (AMDL.US)](https://longbridge.com/zh-HK/quote/AMDL.US.md) - [T-Rex 2X Long ARM Daily Target ETF (ARMU.US)](https://longbridge.com/zh-HK/quote/ARMU.US.md) - [iShares Semiconductor ETF (SOXX.US)](https://longbridge.com/zh-HK/quote/SOXX.US.md) - [VanEck Semiconductor ETF (SMH.US)](https://longbridge.com/zh-HK/quote/SMH.US.md) - [Invesco Semiconductors ETF (PSI.US)](https://longbridge.com/zh-HK/quote/PSI.US.md) - [SPDR S&P Semicon (XSD.US)](https://longbridge.com/zh-HK/quote/XSD.US.md) - [Leverage Shares 2X Long ARM Daily ETF (ARMG.US)](https://longbridge.com/zh-HK/quote/ARMG.US.md) - [AMD (AMD.US)](https://longbridge.com/zh-HK/quote/AMD.US.md) ## 相關資訊與研究 - [Arm Expands Compute Platform to Silicon Products in Historic Company First | ARM Stock News](https://longbridge.com/zh-HK/news/280350392.md) - [Arm Stock Jumps on HSBC Upgrade -- Sees Massive Upside on AI Chip Growth](https://longbridge.com/zh-HK/news/279979761.md) - [Synopsys Supports New Arm AGI CPU with Full-Stack Design Solutions | SNPS Stock News](https://longbridge.com/zh-HK/news/280352334.md) - [Codenotary Introduces First Autonomous Agentic Platform for Securing and Managing Linux Infrastructure and Securing Code](https://longbridge.com/zh-HK/news/280340172.md) - [AMD Eyes Major AI Win with 10,000-Chip Deal in Korea](https://longbridge.com/zh-HK/news/280108451.md)