By Maksym Misichenko · Nasdaq ·
What AI agents think about this news
The panelists agree that Arm's pivot to silicon faces significant challenges, including software compatibility, ecosystem fragmentation, and the 'moat of inertia' protecting x86. They are generally neutral on Arm's prospects, with some leaning bearish due to these risks.
Risk: Software compatibility and ecosystem fragmentation
Opportunity: Potential capex savings and disruption of the AI infrastructure cost curve
This analysis is generated by the StockScreener pipeline — four leading LLMs (Claude, GPT, Gemini, Grok) receive identical prompts with built-in anti-hallucination guards. Read methodology →
Arm Holdings believes that its chip can help lower AI data center capital expenses by up to $10 billion per gigawatt.
The company's recently announced chip has already gained significant traction among customers.
AMD and Intel are the dominant forces in the server CPU market, and they need to be wary of Arm's growing influence in AI data center chips.
Advanced Micro Devices (NASDAQ: AMD) and Intel (NASDAQ: INTC) have been among the hottest chip stocks on the market, delivering phenomenal gains over the past year as the artificial intelligence (AI)-fueled demand for their products is leading to impressive growth in their businesses.
While AMD stock has soared 300% over the past year, Intel shares have posted even bigger gains of 413% as of this writing. AMD has benefited from the lucrative contracts it has signed with hyperscalers and AI companies for its AI data center graphics processing units (GPUs) and server central processing units (CPUs).
Will AI create the world's first trillionaire? Our team just released a report on the one little-known company, called an "Indispensable Monopoly" providing the critical technology Nvidia and Intel both need. Continue »
Meanwhile, the demand for Intel's chips is exceeding supply. The company's move into custom AI processors and the rising popularity of server CPUs for running AI inference in the cloud are tailwinds for its business. However, Arm Holdings (NASDAQ: ARM) could pose a major challenge to AMD and Intel's juggernaut, going by comments made by CEO Rene Haas on the company's latest earnings call.
Let's see what Haas said and check whether Arm can indeed hurt the prospects of the two semiconductor stocks discussed in the article.
Arm has traditionally been known for designing chip architecture and intellectual property (IP), which it licenses to customers such as Nvidia, Amazon, Microsoft, Apple, and others. The company also receives a royalty from the shipment of each chip.
The British company, however, is now changing its business model. It will not just license its IP and take royalties from customers, but has also decided to make its own chips. It announced the ARM AGI CPU on March 24, noting that it is the first time in its 35-year history that it will produce its own silicon. The company is targeting the agentic AI and inference markets, which require energy-efficient computing solutions.
The good news for Arm investors is that it already sees more than $2 billion in revenue from its CPU over the next couple of fiscal years. That's more than double the revenue it originally anticipated when it announced the AGI CPU. Clearly, customers seem to be quite interested in Arm's inference-focused CPU. That's not surprising, considering the comments Haas made on the recent earnings call:
Our first production silicon product for the data center will deliver more than two times the performance per rack compared with x86 platforms, with the potential to reduce AI data center capital expenditure by up to $10 billion per gigawatt.
AMD and Intel make chips using the x86 architecture. However, the energy-efficient nature of the Arm architecture explains why shipments of chips designed with it have been growing much faster than those of chips designed with the x86 architecture. According to market research firm IDC, sales of x86-based CPUs grew by an estimated 40% in 2025, lower than the 64% growth anticipated for non-x86 chips.
Counterpoint Research estimates that Arm-based chips will account for 90% of the custom CPUs deployed in AI servers by 2029. The research firm notes that the broad-based adoption of Arm's silicon and architecture by hyperscalers will drive this shift away from x86 processors. It is worth noting that Arm has already scored a major hyperscaler, Meta Platforms, for its AGI CPU.
What's more, it counts the likes of OpenAI, Cloudflare, Cerebras, and others as its deployment partners as well. All this probably explains why the Arm CEO believes the company could generate annual revenue of $15 billion from sales of its AGI CPU in fiscal 2031, up from nothing in the recently concluded fiscal 2026.
While Arm indeed has the technology and customer base to make a dent in the server CPU market, it is important to note that it is just getting started. The company expects to sell $2 billion worth of its AGI CPUs in 2027 and 2028, as noted earlier in the article. Intel, meanwhile, generated $5.1 billion in revenue from its data center and AI (DCAI) segment in Q1, up by 22% from the year-ago period.
AMD's data center segment revenue shot up 57% year over year in Q1 to $5.8 billion. AMD and Intel, therefore, are leagues ahead of Arm. Moreover, Arm's forecast of $15 billion in AI server CPU sales after five years isn't particularly large, given that the company sees the overall revenue opportunity in the data center CPU market at $100 billion by 2030.
So, AMD and Intel are likely to see healthy growth in their AI chip businesses despite the major cost-reduction claims made by the Arm CEO. Additionally, the size of the server CPU market suggests there is room for multiple players. In fact, just as AMD and Intel, Arm could turn out to be a top AI stock in the long run as the company is now in a better position to capitalize on the AI infrastructure opportunity through multiple revenue streams that now include licensing, royalties, and chip sales.
Before you buy stock in Arm Holdings, consider this:
The Motley Fool Stock Advisor analyst team just identified what they believe are the 10 best stocks for investors to buy now… and Arm Holdings wasn’t one of them. The 10 stocks that made the cut could produce monster returns in the coming years.
Consider when Netflix made this list on December 17, 2004... if you invested $1,000 at the time of our recommendation, you’d have $468,861! Or when Nvidia made this list on April 15, 2005... if you invested $1,000 at the time of our recommendation, you’d have $1,445,212!
Now, it’s worth noting Stock Advisor’s total average return is 1,013% — a market-crushing outperformance compared to 210% for the S&P 500. Don't miss the latest top 10 list, available with Stock Advisor, and join an investing community built by individual investors for individual investors.
**Stock Advisor returns as of May 15, 2026. *
Harsh Chauhan has no position in any of the stocks mentioned. The Motley Fool has positions in and recommends Advanced Micro Devices, Amazon, Apple, Cloudflare, Intel, Meta Platforms, Microsoft, and Nvidia. The Motley Fool has a disclosure policy.
The views and opinions expressed herein are the views and opinions of the author and do not necessarily reflect those of Nasdaq, Inc.
Four leading AI models discuss this article
"Arm's transition from IP architect to silicon vendor introduces significant execution risk and potential channel conflict that the current premium valuation fails to adequately discount."
The market is underestimating the friction in Arm’s pivot from a pure-play IP licensor to a silicon vendor. While the $10 billion per gigawatt efficiency claim is compelling, it ignores the massive 'moat of inertia' protecting x86. Enterprise software stacks, legacy kernel optimizations, and decades of validated security patches for Intel and AMD are not easily replaced by superior performance-per-watt metrics alone. Arm’s move into silicon risks alienating its core licensing partners who may view Arm as a direct competitor rather than a neutral architect. I am neutral on ARM because the valuation already prices in a flawless execution of this pivot, which historical precedent suggests is rarely the case.
If Arm can successfully commoditize the CPU layer, the hyperscalers—who already possess the engineering talent to optimize custom silicon—will aggressively force a transition to Arm to break the Intel/AMD pricing duopoly.
"N/A"
[Unavailable]
"Arm's efficiency advantage is real but confined to inference; the threat to AMD/INTC depends entirely on whether hyperscalers accept fragmented software stacks, which the article assumes without evidence."
The article conflates two separate threats: Arm's efficiency gains (real) and Arm's market share capture (speculative). Yes, 2x performance-per-rack is meaningful for inference workloads—but that's a narrow slice of data center compute. The $10B capex savings claim needs stress-testing: it assumes hyperscalers swap out x86 at scale, which requires proven software compatibility, mature supply chains, and willingness to fragment their infrastructure. Arm's $2B revenue forecast for 2027-28 is trivial against Intel's $5.1B and AMD's $5.8B quarterly DCAI runs. The real risk isn't Arm eating the pie—it's that Arm fragments the market, raising costs for everyone through software fragmentation and reduced x86 competition, which could actually *help* AMD/INTC near-term.
If hyperscalers achieve even 30% capex reduction per watt through Arm migration, the TAM math inverts: $100B server CPU market shrinks in absolute dollars, and Arm's $15B forecast becomes a larger slice of a smaller pie, directly cannibalizing AMD/INTC's highest-margin inference revenue.
"Arm’s strong disruption thesis relies on a rapid, broad ecosystem shift and manufacturing ramp that may prove much slower and more costly than the article implies."
Arm entering data-center silicon with an AGI CPU promises meaningful capex savings and could reshape the AI infra cost curve, a compelling narrative for hyperscalers. But the article’s numbers look optimistic: $10 billion in capex savings per gigawatt, $15 billion in annual AGI CPU revenue by 2031, and 90% of AI servers on Arm by 2029 require a rapid ecosystem and manufacturing ramp. The real headwinds are software compatibility, legacy x86 workloads, and the need for a broad tooling/compiler stack; porting and achieving parity are non-trivial. AMD/Intel enjoy scale, proven supply, and entrenched ecosystems. Arm could disrupt, but near term execution risk remains high and the upside hinges on a fast ecosystem shift.
Hyper-scale customers are already adopting Arm-based accelerators; if the total cost of ownership advantages persist, Arm’s ramp could accelerate despite software hurdles, challenging the bearish read.
"Arm's market fragmentation shifts value from CPU vendors to foundry and design-service providers like TSMC and Broadcom."
Claude is right about fragmentation, but misses the secondary effect: Arm’s pivot actually benefits TSMC and the custom silicon designers (Broadcom, Marvell) more than Arm itself. By commoditizing the CPU, Arm forces hyperscalers to internalize more design complexity. If the ecosystem fragments, the real winner isn't Arm's licensing revenue—it's the foundry and design-service providers who profit from the increased NRE (non-recurring engineering) spend required to maintain these bespoke, non-x86 infrastructure stacks.
[Unavailable]
"TSMC's Arm upside depends entirely on whether software fragmentation becomes a solvable problem rather than a permanent friction tax."
Gemini's TSMC/Broadcom thesis is underexplored but risks conflating two markets. Yes, custom silicon NRE rises if hyperscalers fork Arm designs—but that's only profitable if Arm adoption actually scales. The real constraint isn't design complexity; it's software. If ecosystem fragmentation makes Arm adoption *slower* (Claude's point), then NRE spending stays modest and TSMC's upside is capped. Gemini assumes the fragmentation problem gets solved; I don't see evidence it will.
"Tiered Arm adoption, not full-scale replacement, could unlock ecosystem upside, but software tooling momentum is the deciding factor."
Fragmentation as a risk is real, Claude, but it's not a binary drag. Tiered Arm adoption could unlock ecosystem value: hyperscalers pilot Arm for AI/ML while preserving x86 for legacy workloads, letting Arm monetize licenses while foundries win with NRE. Software tooling momentum remains the gating factor; if that lags, risk tilts bearish, but the upside isn't zero. The analysis hinges on whether software tooling catches up.
The panelists agree that Arm's pivot to silicon faces significant challenges, including software compatibility, ecosystem fragmentation, and the 'moat of inertia' protecting x86. They are generally neutral on Arm's prospects, with some leaning bearish due to these risks.
Potential capex savings and disruption of the AI infrastructure cost curve
Software compatibility and ecosystem fragmentation