Arm’s earnings call resulted in a sharp reversal as the stock was originally up 7%+ after hours yet settled down 6.4%. Overall, management’s commentary raised doubts in two areas – the first is whether Arm can secure the supply to meet AGI CPU demand, and the second is whether Arm has immediate inroads into Big Tech/Big Silicon given these larger players already license Arm’s IP and design their own custom CPUs. If the latter is true, then Arm’s near-term merchant silicon market is enterprises rather than hyperscalers.
Regarding supply, management stated the following on the earnings call: “As Rene noted or as Rene mentioned, customer demand for the ARM AGI CP was very strong. We now have line of sight to more than $2 billion of demand across fiscal '27 and '28. However, we are maintaining our outlook of $1 billion while we pursue supply chain capacity, and we still expect the first revenues from production ship sales to land in the fourth quarter of this fiscal year.”
In other words, regardless of demand, Arm's near-term AGI CPU revenue is supply-capped. The very catalyst that drove the stock's run into the print, anticipated demand for the new CPU, was effectively walked back in the near term.
AGI CPU Demand Doubles; But Won’t See Meaningful Revenue for 1-2 Years
We covered the launch of Arm’s AGI CPU and increasing CPU requirements being driven by agentic AI in the free newsletter, Arm Stock Could Win as Agentic AI Shifts the Bottleneck to CPUs.
Here is what we stated in our previous write-up:
“In agentic workflows, the GPU still handles inference, but between each inference call, the CPU is doing the orchestration – which are best described as handling tool calls, API requests and memory tasks. AI agents are surfacing this new constraint, which is how to prevent latency and underutilized GPUs following the exponential growth of orchestration needs.
For investors, what matters is that CPUs account for 50% to 90% of total latency in workflows, which means the CPU-to-GPU ratio in AI clusters will need to increase. Earlier this year, both AMD and Intel saw analyst upgrades based on the outstripped supply of CPUs leading to higher average sales prices of roughly 10% to 15%. Reuters also reported that Intel’s unfulfilled orders are reaching longer than six months while AMD delivery times are believed to be eight to 10 weeks.outstripped supply of CPUs leading to higher average sales prices of roughly 10% to 15%. Reuters also reported that Intel’s unfulfilled orders are reaching longer than six months while AMD delivery times are believed to be eight to 10 weeks.
Regarding how Arm fits in, the company’s expertise in lowering power requirements could matter more than the market expects. After years of supplying the architecture IP behind other companies’ CPUs, Arm is preparing to directly compete with its customers and x86 CPU competitors by transitioning to a chip designer themselves. This comes during a time when CPU cores are expected to go up 4X from 30 million CPU cores per gigawatt to 120 million CPU cores per GW.”
This long-term agentic AI-driven growth underpins Arm’s shift to chip design with the AGI CPU, with management emphasizing in FQ4 that demand for the AGI CPU is accelerating, with visibility into $2 billion through FY27 and FY28, double what was stated just six weeks ago at the CPU reveal in late March.
Despite this sharp increase in demand visibility, Arm is not yet ready to move the needle for AGI CPU revenue contribution, opting to maintain its FY27-FY28 target at $1 billion as it works to secure more supply. To secure this supply, Arm called out the following: “So the number that we talked about at the end of March was supply in place to support $1 billion of demand. And that includes memory that includes wafers, that includes packaging, that includes access to test equipment. So for the $2 billion, we are now in the process of securing supply to support that.”
To further complicate the supply-capped thesis, TSM recently exited its Arm position.
AGI CPU to Report $1B in Revenue in 2027-2028
Despite the excitement around Arm bringing to market a CPU with 2x the performance of x86 CPUs, the reality is that it will take time to secure memory wafers and ship in volume. The CFO indicated it would be FY28 (aligned to mid-CY2027 to mid-CY2028) before the $1 billion is recognized:
“The revenue split for '27, '28, something like [$90 to $100] million for Q4 '27 and then $910 million or whatever for '28. That's kind of what we laid out 5 or 6 weeks ago. And as said, we have demand above that. But for right now, let's just assume that's the number until we work through some of the wafer memory shortage issues.”
This is a rather long ramp for the AGI CPU, with Arm stating the first revenue will not come until Q4 FY27 (next March) and contribute roughly $100 million – if you wanted to put this in perspective, Nvidia delivered nearly $700 million in data center revenue daily last quarter, and will scale much larger by this time next year.
However, for Arm, the new AGI CPU represents a rather lucrative revenue stream layering in on top of its existing IP business, despite operating far below the scale of Nvidia and AMD. Based on current consensus estimates for $7.61 billion in revenue in FY28, Arm’s projection for ~$910 million in revenue contribution from the AGI CPU would represent nearly 12% of revenue.
Assuming the full $2 billion of demand materializes and 80% converts to revenue in FY28, given Arm has >4 quarters to smooth out the supply chain and secure supply, this may present an ~$800 million uplift to consensus, or >10% on top of the $7.61 billion.
Arm IP Continues to be the Main Growth Driver
Data center royalty more than doubled YoY and is expected to double again in FY27, driven by large customers such as Amazon’s Graviton, Google’s Axion, and Nvidia’s Vera. Arm stated they have over 50% share with top hyperscalers and 100% share in DPUs/SmartNICs: “Royalty revenue grew 11% to $671 million with growth across Edge AI, physical AI and cloud AI, where our data center royalty has more than doubled year-over-year.”
Due to the Arm powering custom CPU programs, the CEO stated he foresees Arm being the largest CPU architecture by the end of the decade: “So we think it's a market that we can play in, in a very large way. And I think even indicators of AWS selling Graviton to outside partners — it's kind of an indication that there's just huge, huge demand for ARM-based capacity. So we think we're going to play alongside our partners in this space. And we also think the opportunity is very, very large for both. And I'm actually confident that by the end of the decade, I believe the largest market share by CPU type will be ARM.”
AMD Flexes Muscle for 50% Market Share with $100-120B TAM; Arm Offers a Rebuttal
AMD set the stage for strong server CPU growth earlier this week as it doubled its long-term industry growth forecast from 18% over the next three to five years to 35%, driven by increasing CPU requirements for agentic AI. This updated forecast now projects the server CPU TAM to reach over $120 billion by 2030, notably 20% higher than the $100 billion TAM Arm forecasted during its AGI CPU launch.
While discussing their new accelerated TAM, AMD’s management mentioned that they are confident in growing to >50% market share, implying a goal of capturing as much as $60 billion of the server CPU market.
In sharp contrast, Arm has stuck to its $15 billion AGI CPU revenue target by FY31, essentially implying a ~12% share based on AMD’s updated $120 billion TAM. Put differently, AMD is aiming to be 4X larger than Arm in AI-driven server CPU revenue by the turn of the decade, presenting stiff competition in server CPUs (Intel isn’t to be forgotten either).
However, when asked on the call about x86, Arm’s CEO offered a controversial take, which is that Arm CPUs will see nearly a 100% attach rate. The original statement was: “Those all connect to Arm. And increasingly, they are going to be 100% Arm. So we feel very, very good about the market share there.”
Here was the question on the call that offered a sharp rebuttal to AMD’s x86 bullish forecast:
“Timm Schulze-Melander Rothschild & Co Redburn
So Rene, maybe just to start with you and to key off that CPU TAM commentary you just made there. I just want to check that I heard you right that you anticipate 100% attach rate of Arm CPU with those accelerators you mentioned? And then maybe just looking forward from an OpEx perspective, as you get into that merchant market, as your products attach to some of your partners' products, do you have any undertakings in terms of operating expenses in terms of in-market customer support? And then I had a quick follow-up for Jason.
Rene Haas CEO & Director
Yes. Thank you for the question, Timm. Yes, so to clarify my comment, my expectation is that for the training platform over time, TPUs over time and NVIDIA's accelerated over time, I believe that the vast majority of the market share there will be Arm. NVIDIA is there essentially, and we are starting to see that happen with Graviton already over the last number of quarters and the announcement that Google made at Google Next with the TPU 8t and 8i, the training and inference chips. So that trend is well underway. And the reason for it, as stated, is that by getting much better performance in the same power envelope, the overall performance of the platform has greatly improved.
Google is talking about an 80% improvement in terms of the overall performance. So it's really numbers like that and the advantages that customers see in terms of embracing the platform that gives us very, very high confidence that, that trend should continue […]”
Technically, statements from both management teams offer an element of truth as there are many x86-hosted AI accelerator servers shipping today (hence Intel’s and AMD’s strong reports). While many hyperscalers deploy Arm-based servers internally, those same hyperscalers still run substantial x86 capacity for customer workloads.
Net-net, AI servers are primarily x86-hosted whereas mobile is entirely Arm-hosted. Arm is betting on a massive shift in the coming years, whereas AMD is offering the incumbent’s view.
In an important exchange with Vivek Arya, Arm CEO Rene Haas admitted the numbers don’t add up when taking management projections at face value for year-end CPU market share from AMD, Intel and Arm: “As far as the market share numbers, AMD has 50, Intel has 50 and we have 50. So you add up to some crazy number.”
Haas also had a quick comment about the AGI CPU’s initial customers that carries quite an important readthrough. Launch partners like Cloudflare, SAP or SK Telecom are adopting the chip because they do not have the capex budgets and/or engineering expertise to design and deploy custom Arm-IP based CPUs at scale – this will likely remain at the hyperscaler level with chips such as Amazon’s Graviton or Google’s Axion.
The main readthrough from Arm’s answer here is that will primarily be serving the enterprise market with the AGI CPU, having to compete with AMD and others for customers wanting internal CPU capacity, while also having to make a compelling argument for customers to adopt the chip instead of simply using Arm-based CPUs like Graviton in the cloud. It also hints that there may not be much of a runway for the AGI CPU at the hyperscalers who do indeed have the budgets and have already successfully deployed Arm-based custom CPUs at scale.
The truth is that – nobody knows how this will play out exactly. AMD’s management team doubled their TAM very quickly in a way that suggests they were caught off guard by the demand signals. Therefore, long-term forecasts are hard to predict in this space.
Financials:
Q4 Revenue Grew by 20%
Arm’s Q4 FY26 revenue grew by 20% YoY and 20% QoQ to a record $1.49 billion, beating the midpoint of management’s guidance ($1.470 billion) by 1.36%.
Royalty revenue decelerated from 27% YoY in Q3 to 11% YoY in Q4 with revenue of $671 million; this also represented a (9%) QoQ decline off Q3’s strong $737 million. YoY growth was driven primarily Cloud AI with data center royalties more than doubling YoY. Arm also continues to benefit from an increasing mix shift to Armv9 and CSS, which carry meaningfully higher per-chip royalty rates than prior architectures.
License and other revenue grew 29% YoY and 62% QoQ to $819 million, driven by continued strong demand for Arm IP, the timing and size of multiple high-value license agreements and contributions from backlog.
Management guided Q1 FY27 revenue to $1.26 billion at the midpoint (+/- $50 million), implying YoY growth of 19.7% but down (15.4%) QoQ on the typical seasonality that follows a Q4 license catch-up. The Q1 guide is roughly in line with consensus of $1.25 billion. Both royalty revenue and license and other revenue were guided to be up around 20% YoY in Q1 FY27.
For the full year, FY26 revenue grew 23% YoY to a record $4.92 billion — the third consecutive year of more than 20% revenue growth since IPO — with royalty revenue up 21% YoY to $2.61 billion and license revenue up 25% YoY to $2.31 billion. Looking ahead, analysts expect FY27 revenue to decelerate slightly to 20.9% YoY to $5.92 billion, before reaccelerating to 28.5% YoY to $7.61 billion in FY28, the latter benefitting from initial contribution of the Arm AGI CPU silicon business.
ACV Growth Decelerates to 22% YoY
Annualized contract value (ACV), management’s preferred metric for normalized license and other revenue, grew 22% YoY and 2% QoQ to $1.66 billion; this marked a six point deceleration from 28% YoY growth maintained over the last three quarters.
Remaining performance obligations (RPO), however, declined (7%) YoY and (4%) QoQ to $2.07 billion, marking the third consecutive quarter of YoY RPO declines, which management attributed to improvements in the timing of revenue conversion (i.e. faster recognition rather than weakening demand).
Arm also signed two more CSS licenses in the quarter, one for smartphones and the other for data center networking chips. Arm Total Access licenses increased by 6 in the quarter to 56 (up 27% YoY), now including more than half of Arm’s top 30 customers, while Arm Flexible Access customers increased by 11 to 329 (up 5% YoY).
Margins
Gross margin remained near best-in-class IP-business levels, but operating margin compressed YoY (despite opex coming in below guidance) as Arm continued to invest in R&D for the AGI CPU and CSS roadmaps. Management has indicated that FY26 should mark the peak of opex growth, with non-GAAP opex CAGR decelerating from a 26% pace in FY24-FY26 to a mid-teens CAGR through FY31, which should drive operating leverage.
- Q4 GAAP gross margin was 97.9%, up slightly from 97.7% a year ago. Non-GAAP gross margin was 98.3%, essentially flat with 98.4% a year ago.
- Q4 GAAP operating margin was 29.4%, down from 33.0% in the prior year period. Q4 adjusted operating margin was 49.1%, down from 52.8% a year ago, as adjusted operating expenses grew ten points faster than revenue, up 30% YoY to $734 million
- Q4 GAAP net margin was 21.0%, up from 16.9% a year ago, while adjusted net margin of 43.0% declined from 47.1% a year ago.
Full-year FY26 GAAP and non-GAAP gross margins were 97.5% and 98.2%, respectively, both increasing roughly half a point YoY. However, operating margins felt some pressure, with FY26 GAAP operating margin contracting 2.4 points to 18.3%, and adjusted operating margin contracting 3.7 points to 43%. This was driven by strong opex growth, up 33% YoY to $2.72 billion, 13 points faster than revenue.
This operating margin contraction flowed through to the bottom line, with FY26 GAAP net margin of 18.4%, down 1.4 points, and adjusted net margin of 38.4%, down nearly 5 points.
Adjusted EPS Grew 9%
Q4 adjusted EPS was $0.60, up 9.1% YoY and beating estimates for $0.58. GAAP EPS in the quarter was $0.29, up 45% YoY but missing consensus of $0.37 by (20.7%) on the higher GAAP opex line. Management guided Q1 FY27 non-GAAP fully diluted EPS to $0.40 at the midpoint (+/- $0.04), implying 12.5% YoY growth.
For the full year, FY26 adjusted EPS was a record $1.77, up 8.6% YoY, while GAAP EPS increased 13.3% to $0.85. Analysts expect FY27 adjusted EPS to accelerate to 21.4% YoY to $2.14 (up 21.4% YoY), with this accelerating extending further into FY28, up 37.4% YoY to $2.94.
Cash Flow and Balance Sheet
Cash flow generation was strong on a full-year basis, although Q4 cash conversion was light.
- Q4 operating cash flow was $260 million, essentially flat with $258 million a year ago, for an operating cash flow margin of 17.4%, down from 20.8% a year ago as receivables grew. FY26 operating cash flow was $1.52 billion for a 31% margin, up sharply from FY25’s $397 million for a 9.9% margin.
- Q4 adjusted free cash flow was $152 million, down from $163 million a year ago, for an FCF margin of 10.2%, down from 13.1%. FY26 adjusted free cash flow was $882 million for a 17.9% margin, up from just $99 million in FY25 (a 2.5% margin).
- Cash and short-term investments totaled $3.60 billion at quarter-end, up from $3.54 billion in Q3, and the company continues to carry no debt.
Conclusion:
My view is that Arm remains a critical long-term player in the AI data center buildout, but this earnings report introduced more uncertainty around the near-term growth story. The issue is less about AMD or Intel’s current dominance in x86 and more about Arm’s ability to secure supply at a time when even the largest AI semiconductor companies are capacity constrained.
After years of appearing relatively uneventful compared to other AI semiconductor peers, Arm is now better positioned to compete as AI workloads expand from mobile and edge devices into the data center. The key unknown is valuation, especially because merchant CPU revenue will take time to scale, and investors may need to rely on Arm’s traditional IP engine as the primary growth lever for the next one to two years.
My takeaway is that the near-term AGI CPU narrative should be priced lower due to a capped growth trajectory, but Arm’s long-term strategic relevance remains intact.
Please note: The I/O Fund conducts research and draws conclusions for the company’s portfolio. We then share that information with our readers and offer real-time trade notifications. This is not a guarantee of a stock’s performance and it is not financial advice. Please consult your personal financial advisor before buying any stock in the companies mentioned in this analysis.
Recommended Reading: