ARM Holdings Shares Surge 15% on News of Partnership with OpenAI
British chip developer signs agreement to create a specialized AI processor. Market cap on Nasdaq exceeds $140 billion.
Analytical Breakdown: ARM's 15% Surge — The Architectural Revolution You Missed
[The Gist]: What's Really Happening
Headlines scream about "ARM's partnership with OpenAI," but the real reason for ARM's 15% stock surge is entirely different — and far bigger than a single contract with Sam Altman. What happened on June 1, 2026, at the Computex trade show in Taipei is a tectonic shift in the semiconductor industry, comparable only to the launch of the iPhone in 2007. NVIDIA officially announced the creation of the RTX Spark — the first system-on-a-chip for Windows laptops and compact desktops built on the ARM architecture.
This isn't just "another chip." The RTX Spark is a superchip combining a 20-core ARM processor (co-developed with MediaTek), a Blackwell GPU with 6,144 CUDA cores, up to 128 GB of unified memory via NVLink, and 1 petaflop of performance for AI tasks. Manufacturing uses TSMC's 3-nanometer process. In fall 2026, over 30 laptop models and 10 desktop models from Dell, HP, Lenovo, ASUS, and MSI will hit the market with this chip inside.
And here's the core point that 99% of commentators miss. ARM shares surged 15% not because OpenAI will use their chips, but because NVIDIA has just legitimized ARM as the only viable alternative to x86 in the personal computer segment. Apple did this for itself with the M-series, but NVIDIA is doing it for the entire industry. Intel and AMD held a duopoly in the PC processor market for decades. As of today, that duopoly is dead.
Timeline and Context
The path to this moment took years, and it's important to understand each step. On May 20, 2026, NVIDIA reported Q1 of fiscal 2027 with revenue of $81.6 billion, up 85% year-over-year [from previous analysis]. NVIDIA shares rose, but not dramatically — about 1.3% after the report. However, ARM surged 15% the same day. Why the difference? Because investors understood: NVIDIA needs ARM, and this need will only grow.
On May 29-30, 2026 — two days before the announcement — ARM insiders were actively selling shares. According to SEC filings, on May 20-21, ARM top executives sold over $15 million worth of stock: Jason Child ($7.2 million), Richard Grisenthwaite ($5.1 million), Charlotte Eaton ($2.3 million), William Abbey ($1.3 million), and others. This is a classic pattern: insiders lock in profits before a major pullback or, conversely, before news they've already priced in. In this case, it's the latter. They knew about Computex.
On June 1, 2026, at 10:00 AM Taipei time (10:00 PM ET on May 31), Jensen Huang takes the stage. Two major announcements: RTX Spark (PC chip on ARM) and confirmation that the Vera Rubin platform (Vera CPU + Rubin GPU) for data centers has entered mass production. OpenAI, Anthropic, xAI, Dell, Oracle, and CoreWeave are already customers of Vera Rubin.
Market reaction is immediate and brutal. ARM surges 15%. NVIDIA rises 6.3%. Intel falls 5-6%, AMD 4-5%, Qualcomm 8-9%. Dell and HP rise 4-9% as launch partners. ARM's market cap exceeds $140 billion, closing June 1 at around $353 per ADR.
But the key detail almost no one noticed: the Vera Rubin announcement (Vera CPU + Rubin GPU) means NVIDIA is now producing a full-fledged server CPU on the ARM architecture for data centers. This is a direct invasion of Intel and AMD's most profitable segment. And RTX Spark finishes them off in the consumer segment. NVIDIA is attacking on all fronts simultaneously.
Who Wins and Who Loses
Winner #1: ARM Holdings. This is obvious, but the scale is not. ARM has a problem no one talks about: SoftBank owns 86.72% of the company, with a free float of only 14%. This means every dollar of institutional demand creates 7 times more price movement than in a normal company. Short interest as of May 31 was over 16 million shares. When a positive catalyst like the RTX Spark announcement hits, short sellers are forced to cover, adding another 30-40% momentum to the rally.
But this is a double-edged sword. With a P/E ratio of about 415 and a P/B of about 45, ARM shares are already trading as if the company will conquer the world tomorrow. Any delay in royalties from NVIDIA or other licensees will cause a drop 2-3 times more severe than a normal stock. Citi has already raised its target to $300 (from the current $353, that would be a drop if the target is below the market?) — Wait, there's a discrepancy. The current price is around $353, and Citi's target is $300. This means even Citi considers the stock overvalued by 15% after the rally.
Winner #2: NVIDIA. The RTX Spark is not just a new product. It completes the puzzle "from data center to pocket." NVIDIA can now offer a customer an ARM chip for a phone (via MediaTek), a laptop (RTX Spark), a server (Vera CPU), and a supercomputer (Rubin GPU). No one in the world has such a complete stack. The real win is locking developers into the CUDA ecosystem. If all AI code is written in CUDA, and CUDA only runs on NVIDIA + ARM, then Intel and AMD lose not just chip sales — they lose the chance to participate in the future of computing.
Winner #3: MediaTek. The Taiwanese semiconductor manufacturer co-developed the ARM core for the RTX Spark with NVIDIA. The media barely covers this, but it's a colossal breakthrough for MediaTek. They were long seen as a "chipmaker for budget phones," and now they're involved in creating a premium AI PC chip alongside NVIDIA. MediaTek shares on the Taiwan Stock Exchange surged 5% on the day of the announcement.
Loser #1: Intel. This is a catastrophe for Intel on the scale of 2006, when AMD introduced the K8 architecture. Intel dominated the PC processor segment for decades thanks to the x86 architecture and its tie to Windows. Now NVIDIA + Microsoft (Windows on ARM) + OEM partners (Dell, HP, Lenovo) are creating an alternative ecosystem that could capture 20-30% of the market in 2-3 years. Intel's 5-6% drop on the day of the announcement is just the beginning. I expect Intel to trade in the $18-22 range over the next 6 months, 30-40% below current levels.
Loser #2: AMD. AMD is in a more complex position than Intel. AMD has good products (Ryzen, EPYC), but they're all on x86. In data centers, AMD tries to compete with NVIDIA using Instinct GPUs, but the software gap (CUDA vs. ROCm) remains enormous. Now NVIDIA is also entering the PC segment with an ARM chip that, according to preliminary tests, outperforms Ryzen AI in performance per watt. AMD fell 4-5%, but in my estimation, there's another 10-15% downside by the end of summer.
Loser #3: Qualcomm. This is the most underappreciated loser. Qualcomm spent years developing the Snapdragon X Elite — a platform for Windows on ARM, trying to become the "ARM processor for PCs." NVIDIA just announced it will compete with Qualcomm on their own turf, but with an advantage: NVIDIA has a GPU, which Qualcomm lacks, and NVIDIA has ties with OEMs that are stronger than Qualcomm's. The Snapdragon X Elite now looks like a "Nintendo vs. PlayStation" — a losing battle from the start. Qualcomm fell 8-9%, and I see no reason for recovery in the coming quarters.
What the Media Isn't Saying
Non-obvious Insight #1: 14% free float is a ticking time bomb. Let's revisit ARM's capital structure. SoftBank owns 86.72% of shares. Institutional investors hold about another 8%. The actual tradable free float is less than 14%. Meanwhile, short interest is 16.1 million shares. This means short interest as a percentage of free float is about 20-25%.
In a normal company, such a level of short positions is a concern. In a company with such low free float, it's dynamite. Any positive catalyst triggers not just a rally, but exponential growth due to mass short covering. The 15% rise on June 1 is just the tip of the iceberg. If NASDAQ continues to rise and ARM gets another positive catalyst (e.g., confirmation that Vera Rubin receives orders from Amazon or Google), we could see another 20-30% gain in 2-3 days. But if negative news hits, the drop will be just as explosive.
Non-obvious Insight #2: Microsoft is a silent beneficiary, and no one is talking about it. Microsoft has an exclusive agreement with ARM and NVIDIA to optimize Windows on ARM for the RTX Spark. This gives Microsoft something it never had: the ability to compete with Apple MacBooks on equal footing in performance and battery life. The Apple M-series was out of reach for Windows laptops for years. The RTX Spark, if specs are to be believed, closes the gap to a minimum.
But more importantly, Microsoft is no longer dependent on Intel. This is enormous bargaining power. If Intel raises processor prices, Microsoft can threaten to switch to ARM. If Qualcomm is too slow in developing Snapdragon, Microsoft can fully pivot to NVIDIA + MediaTek. Microsoft has created insurance against the x86 monopoly, and no one is pricing this into MSFT shares.
Non-obvious Insight #3: The ARM vs. x86 war will be won not in PCs, but in data centers. The RTX Spark is a loud but not the most important announcement. The most important announcement is the mass production launch of Vera Rubin. The Vera CPU, which NVIDIA developed on the ARM architecture, claims 3.5 times higher AI training performance and 5 times higher inference performance compared to the Blackwell platform.
What does this mean? It means the largest hyperscalers (Amazon, Google, Microsoft, Meta) can now build data centers on NVIDIA's Vera processors instead of Intel Xeon or AMD EPYC. In data centers, ARM wins not on low price, but on performance per watt. NVIDIA has just accelerated the transition from x86 to ARM in AI infrastructure by about 2-3 years.
Forecast: Next 30 Days and 90 Days
30 days (through early July 2026): Key event is ARM's Q1 fiscal 2027 earnings report, expected in mid-July. Consensus revenue forecast is $1.255-1.265 billion, growth of about 20% year-over-year. But more important will be guidance: if ARM raises its current quarter forecast due to royalties from NVIDIA and other licensees, shares could continue to $380-400. If rumors that SoftBank is considering selling part of its stake (which would increase free float and reduce volatility) are confirmed, shares could correct to $280-300.
Technical analysis: current price $353, resistance at the all-time high around $356. Support at $310 (50-day moving average). RSI on the daily chart is about 72 — overbought territory, but in a short squeeze, this is not an obstacle. Risk: any negative analyst comment about overvaluation (P/E 415 vs. industry average 35) could trigger profit-taking.
90 days (through September 2026): Here the scenario bifurcates. Base case: ARM reports Q1 in line with forecasts ($1.26 billion revenue), confirms that royalties from NVIDIA RTX Spark will start in Q4 2026, and shares consolidate in the $320-370 range with elevated volatility. Bull case: NVIDIA announces that pre-orders for RTX Spark exceed 10 million chips for 2026-2027, ARM raises its annual royalty forecast by 30-40%, and shares surge to $450-500. Bear case: production delays for RTX Spark due to issues at TSMC, and ARM falls to $250-270.
Structurally, by end of 2026, ARM should trade around $380-420 if the Windows on ARM ecosystem scales successfully. But any investor must understand: ARM shares are currently priced for a perfect scenario. The slightest deviation — and the price will crash 30-40% due to the same low free float that fueled the surge.
Editorial Forecast
Asset: ARM Holdings (NASDAQ: ARM). Direction: Profit-taking and correction within the next 24-72 hours. Key levels: resistance — $356 (all-time high), support — $335 (June 1 close) and then $310. Confidence level: medium (60%). Main risk: institutional investors who missed the rally may start buying on any pullback, limiting the correction's depth. RSI at 72 indicates overbought conditions, and trading volume of 10.45 million shares is significantly above average — signs that the rally was driven by short covering, not a fundamental revaluation. It is recommended to wait for a pullback to $310-320 to open long positions with a 6-12 month horizon.
— Editorial Team
