For a decade, the AI chip conversation had one name: Nvidia. As of early June 2026, it has two. On May 5, 2026, Advanced Micro Devices reported a first-quarter that turned its long-running underdog story into a genuine two-horse race – a record data center business, its highest-ever quarterly profit, and a customer roster that now includes signed, multi-gigawatt commitments from both OpenAI and Meta. The market’s verdict was immediate: AMD stock jumped more than 14% in after-hours trading, briefly trading above $413 a share.
The AMD earnings print landed two weeks before Nvidia’s own blockbuster quarter, and the contrast defined the moment. Nvidia is still an order of magnitude larger. But for the first time since the AI boom began, investors are pricing in a world where a second supplier matters – and in 2026, AMD’s stock has sharply outrun its bigger rival. This is a deep look at the numbers behind the quarter, the two deals reshaping the AI hardware map, the Instinct MI450 roadmap AMD is betting on, and how it all stacks up against Nvidia’s $75.2 billion data center machine.
Don't miss new tech stories on Google
Add Tech Insider once in the Google app and our stories appear in your news suggestions.
AMD Earnings Q1 2026: A Record $5.8B Data Center Quarter
AMD posted total revenue of $10.25 billion for the first quarter of 2026, up 38% year over year and roughly flat sequentially. The engine was the Data Center segment, which delivered a record $5.8 billion – up 57% from a year earlier and, for the first time, more than half of the entire company’s revenue. Data Center operating income reached $1.6 billion, according to figures reported by TECHi and DataCenterDynamics.
Profitability told the more dramatic story. GAAP net income hit $1.4 billion, up 95% year over year – nearly doubling – while non-GAAP earnings per share came in at $1.37, up 43%. GAAP EPS was $0.84. Gross margin expanded to 53% on a GAAP basis and 55% non-GAAP, and free cash flow reached a record $2.57 billion for the quarter. Every headline metric moved in the right direction at once, which is precisely why AMD stock reacted the way it did.
| Metric (Q1 2026) | Result | Year-over-Year |
|---|---|---|
| Total revenue | $10.25 billion | +38% |
| Data Center revenue | $5.8 billion (record) | +57% |
| Data Center operating income | $1.6 billion | – |
| GAAP net income | $1.4 billion | +95% |
| Non-GAAP diluted EPS | $1.37 | +43% |
| GAAP diluted EPS | $0.84 | – |
| Non-GAAP gross margin | 55% | +~2 pts |
| Free cash flow | $2.57 billion (record) | – |
Inside the Segments: Where AMD’s Growth Came From
The segment breakdown shows how lopsided AMD’s business has become toward AI infrastructure. Data Center – which houses both EPYC server CPUs and Instinct AI accelerators – was the clear driver, but the Client segment (desktop and laptop processors) also had a strong quarter, and even the smaller units grew. That balance matters: unlike a pure-play accelerator vendor, AMD’s AMD data center franchise rides on the back of an entrenched server-CPU business that keeps it inside every major cloud data center.
| Segment | Q1 2026 Revenue | YoY Change | Share of Total |
|---|---|---|---|
| Data Center | $5.8 billion | +57% | ~56% |
| Client | $2.89 billion | +26% | ~28% |
| Embedded | $873 million | +6% | ~9% |
| Gaming | $720 million | +11% | ~7% |
| Total | $10.25 billion | +38% | 100% |
Management’s guidance reinforced the trajectory. AMD projected second-quarter 2026 revenue of $11.2 billion, plus or minus $300 million – implying roughly 46% year-over-year growth and 9% sequential growth – with non-GAAP gross margin near 56%. The company said it expects server CPU revenue to grow more than 70% year over year in Q2, and reiterated a longer-term view that the total server CPU market could exceed $120 billion by 2030. That is the quiet, high-margin base beneath the flashier AI numbers.
Why AMD Stock Is Soaring in 2026
The post-earnings pop was not an isolated event – it capped one of the best runs for any large-cap semiconductor stock this cycle. Through 2026, AMD stock has been up roughly 66% year to date, dramatically outpacing Nvidia’s gain of around 5% over the same stretch, as The Motley Fool and other analysts have documented. Some full-year analyses peg AMD’s advance even higher. Whatever the exact figure, the direction is unambiguous: investors have rotated toward the challenger.
Three things changed the narrative. First, the raw AMD earnings momentum – a data center business compounding at nearly 60% with expanding margins. Second, proof of demand from the two most closely watched buyers in AI. Third, a credible next-generation product in the Instinct MI450, arriving exactly when hyperscalers are desperate for a second source. Each on its own would move a stock; together they re-rated the entire company.
The OpenAI Deal: 6 Gigawatts and a 10% Warrant
The turning point arrived in October 2025. AMD and OpenAI announced a strategic partnership to deploy 6 gigawatts of AMD Instinct GPUs across multiple hardware generations. The first 1-gigawatt tranche is scheduled to come online in the second half of 2026 using AMD’s next-generation Instinct MI450 accelerators. To put 6 gigawatts in perspective, that is enough power draw to rival the output of several large nuclear reactors – a staggering commitment of compute.
The structure was as notable as the size. AMD issued OpenAI a warrant for up to 160 million shares of AMD stock – roughly 10% of the company – vesting in tranches as deployment milestones are met and as AMD’s share price hits specified targets, per reporting from CNBC. In effect, AMD aligned its largest new customer as a prospective shareholder, and analysts estimated the arrangement could translate into revenue on the order of $100 billion over its life. It was the clearest signal yet that OpenAI wanted to break its single-vendor dependence.
The Meta Deal: A Second 6GW Mega-Commitment
If the OpenAI deal was a shock, Meta’s was confirmation of a pattern. On February 24, 2026, AMD and Meta announced an expanded partnership to deploy another 6 gigawatts of AMD GPUs across multiple Instinct generations – including a custom accelerator based on the MI450 architecture. Shipments are slated to begin in the second half of 2026, as detailed by ServeTheHome.
AMD deployed a nearly identical warrant to the one it gave OpenAI: a performance-based grant for up to 160 million shares, again roughly 10% of the company, vesting against shipment milestones and share-price thresholds. The announcement added tens of billions of dollars to AMD’s market capitalization in a single session. Two of the five largest AI spenders on earth had now publicly committed to AMD silicon at gigawatt scale – a validation no marketing budget could buy.
| Deal Terms | OpenAI | Meta |
|---|---|---|
| Announced | October 6, 2025 | February 24, 2026 |
| Capacity committed | 6 gigawatts | 6 gigawatts |
| Lead product | Instinct MI450 | Custom MI450-based accelerator |
| First deployment | H2 2026 (1 GW) | H2 2026 |
| Equity warrant | Up to 160M shares (~10%) | Up to 160M shares (~10%) |
| Vesting | Milestones + share-price targets | Milestones + share-price targets |
Instinct MI450 and Helios: AMD’s Answer to Blackwell
None of these commitments matter without competitive silicon, and this is where AMD has closed the most ground. The company unveiled its full MI400 series at its Advancing AI 2025 event and again at CES 2026. The flagship Instinct MI455X is specced at 40 petaflops of FP4 compute, 432 GB of HBM4 memory, and 19.6 TB/s of memory bandwidth – figures that put it squarely in the same conversation as Nvidia’s top parts rather than a generation behind.
The rack is the new unit of compute
AMD’s most important shift is that it is now selling systems, not just chips. Its rack-scale platform, Helios, is designed to deliver up to 3 AI exaflops per rack, directly targeting Nvidia’s rack-scale approach. AMD guided to initial Helios volume in Q3 2026 with a significant ramp across Q4 2026 and into Q1 2027 – the exact window in which the first OpenAI and Meta gigawatts are supposed to light up. The AMD Instinct MI450 is the linchpin: it is sampling with lead customers now, with production shipments ramping in the second half of 2026.
The strategic logic is that hyperscalers no longer buy loose GPUs; they buy integrated racks with networking, memory, and power engineered together. By meeting Nvidia at the rack level with Helios, AMD removes the biggest historical objection to its accelerators – that they were fast chips trapped in an immature systems-and-software stack.
AMD vs Nvidia: The 2026 AI Chip Scoreboard
Context is everything. Two weeks after AMD reported, Nvidia posted first-quarter fiscal 2027 revenue of $81.6 billion, up 85%, with data center revenue of $75.2 billion, up 92%, per its reported results. Nvidia’s data center business alone is roughly thirteen times the size of AMD’s. Nvidia still commands an estimated 80–90% of the AI accelerator market, and its market capitalization sits near $5 trillion, making it the most valuable company on earth.
| Latest Quarter | AMD (Q1 2026) | Nvidia (Q1 FY2027) |
|---|---|---|
| Total revenue | $10.25B | $81.6B |
| Data center revenue | $5.8B | $75.2B |
| Data center YoY growth | +57% | +92% |
| Non-GAAP gross margin | 55% | ~75% |
| Est. AI accelerator share | Low double digits | ~80–90% |
| 2026 stock performance (YTD) | ~+66% | ~+5% |
| Flagship AI part | Instinct MI455X / MI450 | Blackwell (GB300), Rubin next |
The scoreboard reveals the paradox at the center of the trade: Nvidia is winning on every operational metric that matters – scale, growth rate, and a ~75% gross margin that dwarfs AMD’s 55% – yet AMD’s stock has been the far better performer in 2026. The market is pricing the second derivative. Nvidia’s dominance is priced in; AMD’s ascent from a low base is not.
The Warrant Math: What OpenAI and Meta Dilution Means
The most debated feature of the AMD story is those warrants. Combined, OpenAI and Meta could ultimately hold warrants for up to about 320 million AMD shares – a meaningful slice of a company with roughly 1.6 billion shares outstanding. Bulls call it demand you can bank; skeptics call it revenue AMD is partially paying for with its own equity. A simplified way to frame the trade-off:
# Illustrative warrant dilution (not official AMD figures)
shares_outstanding = 1_600_000_000
openai_warrant_max = 160_000_000 # up to ~10%
meta_warrant_max = 160_000_000 # up to ~10%
combined_warrants = openai_warrant_max + meta_warrant_max # 320,000,000
max_dilution_pct = combined_warrants / (shares_outstanding + combined_warrants)
# => ~16.7% fully diluted IF every tranche vests
# But warrants only vest as GPUs ship AND the stock rises.
# Full vesting implies AMD already captured 12 GW of demand
# and a much higher share price – i.e. the "cost" only triggers
# in the scenario where the deals worked spectacularly.That last point is the crux. The warrants are engineered so that maximum dilution only occurs in the world where AMD has already shipped enormous volume and its share price has climbed sharply. In other words, the dilution is self-funding: shareholders only pay it if they have already won. It is an elegant structure, but it does concentrate AMD’s fortunes in a small number of customers – the central risk we return to below.
The $725B Hyperscaler Wave Lifting Both Chipmakers
Neither AMD nor Nvidia operates in a vacuum. The demand under both is a historic build-out of AI infrastructure by a handful of hyperscalers. Google, Amazon, Microsoft, and Meta are collectively on track to spend roughly $725 billion in capital expenditure in 2026, up about 77% from the prior year’s record, according to Tom’s Hardware. Amazon alone is guiding toward roughly $200 billion, Microsoft near $190 billion, Alphabet $175–185 billion, and Meta $115–135 billion.
Analysts estimate around 75% of that spend – roughly $450 billion – funds AI-specific infrastructure: GPUs, custom silicon, data center shells, and power. With projections already floating past $1 trillion in combined capex for 2027, the addressable market is expanding faster than any single vendor can serve it. That is exactly the condition under which a credible second supplier thrives: buyers actively want AMD to succeed, because a monopoly supplier is a supply-chain risk they cannot afford.
Historical Context: From Survival to AI Contender
The scale of this turnaround is hard to overstate. A decade ago, AMD was a near-penny-stock fighting for survival, written off as a permanent also-ran to Intel in CPUs and to Nvidia in graphics. Under CEO Lisa Su, the company rebuilt around the Zen CPU architecture, used its EPYC server chips to claw serious share out of Intel’s data center stronghold, and acquired Xilinx to broaden into adaptive and embedded computing.
What changed in 2025–2026 is that AMD finally translated that CPU credibility into AI accelerator credibility. For years, the knock on AMD was software – that its ROCm stack could not match Nvidia’s CUDA moat. The OpenAI and Meta deals are, implicitly, a statement that AMD’s software and systems have crossed a threshold. When the two companies with the most demanding AI workloads on the planet sign multi-generation commitments, the “software isn’t ready” argument loses much of its force. This is the same competitive reset now visible across the industry, from AMD’s platform roadmap to Intel’s efforts to reinvent its foundry business.
Risks: Concentration, Delivery, and Nvidia’s Moat
The bull case has real holes. The most obvious is customer concentration. A future where OpenAI and Meta each represent a huge share of Instinct demand is a future where the loss – or slippage – of either deal is catastrophic to the thesis. Both commitments are multi-year and milestone-based, meaning most of the promised revenue is not yet contracted, shipped, or paid.
The second risk is execution. Helios and the MI450 ramp are scheduled for the second half of 2026 – an aggressive timeline for a first-generation rack-scale system competing against Nvidia’s mature, sold-out platforms. Any delay pushes the OpenAI and Meta gigawatts to the right and undercuts the growth narrative baked into AMD stock. The third risk is simply Nvidia itself: a ~75% gross margin, a next-generation Rubin architecture ramping in the same H2 2026 window, and a CUDA software ecosystem that remains the industry default. AMD has closed the gap; it has not erased it.
There is also the memory question. The MI455X leans on 432 GB of HBM4, and HBM is the single tightest link in the AI supply chain amid a broader memory shortage that has pushed DRAM and HBM allocation to breaking point. AMD’s ramp depends on securing that memory at volume – a constraint entirely outside its control.
What the Market Is Watching Next
For the rest of 2026, three data points will govern AMD stock. The first is the Q2 print against that $11.2 billion guide – specifically whether Data Center reaccelerates and whether server CPU revenue delivers the promised 70%-plus growth. The second is any hard evidence that Helios and MI450 are shipping on schedule in Q3. The third is the size and cadence of new customer wins beyond OpenAI and Meta; a third or fourth gigawatt-scale name would confirm this is a platform shift, not two idiosyncratic deals.
Investors will also weigh AMD against the broader field. Nvidia’s own record quarter and specialized parts like the RTX Spark superchip keep raising the bar, and the entire sector remains hostage to memory pricing. The question is no longer whether AMD is a real AI player – the AMD earnings answered that – but how large a share of a trillion-dollar build-out it can realistically capture.
Predictions: AMD’s AI Trajectory Through 2027
- Data center crosses the halfway mark for the full year. With Q1 already at ~56% of revenue and MI450 ramping, expect Data Center to remain AMD’s majority segment for all of 2026 and to approach an $8–10 billion quarterly run-rate as the first OpenAI and Meta gigawatts deploy in H2.
- Helios timing is the swing factor. If rack-scale volume genuinely ships in Q3 2026 as guided, AMD’s 2027 estimates get revised up sharply. Any slip into 2027 does the opposite – this is the single most important date on the calendar.
- AMD’s accelerator share rises, but Nvidia stays dominant. Expect AMD to climb into low-double-digit AI accelerator share through 2026 while Nvidia holds above 75%. The market is expanding fast enough for both to grow in absolute terms.
- More warrant-style mega-deals follow. The OpenAI/Meta template – capacity commitment plus equity warrant – is likely to be repeated with additional hyperscalers or sovereign-AI buyers, intensifying the debate over dilution versus demand.
- Volatility stays high. A stock that has roughly doubled on expectations trades on narrative. Each earnings date and shipment milestone will produce outsized moves in either direction.
Frequently Asked Questions
How much did AMD earn in Q1 2026?
AMD reported total revenue of $10.25 billion (up 38% year over year) and GAAP net income of $1.4 billion (up 95%). Non-GAAP EPS was $1.37 and GAAP EPS was $0.84. The standout was a record Data Center segment at $5.8 billion, up 57% and more than half of company revenue.
Why is AMD stock up so much in 2026?
AMD stock has been up roughly 66% year to date in 2026, far outpacing Nvidia’s ~5%, driven by three catalysts: a rapidly growing, high-margin data center business; multi-gigawatt AI GPU commitments from OpenAI and Meta; and a competitive next-generation product in the Instinct MI450. The May earnings beat added a further 14% after-hours pop.
What are the AMD OpenAI and Meta deals?
Both are commitments to deploy 6 gigawatts of AMD Instinct GPUs across multiple generations. OpenAI’s was announced in October 2025 (first gigawatt in H2 2026 on MI450); Meta’s expanded partnership was announced in February 2026 with a custom MI450-based accelerator. In each case AMD issued the customer a warrant for up to 160 million shares – roughly 10% of the company – vesting on shipment and share-price milestones.
How does AMD compare to Nvidia in AI chips?
Nvidia remains far larger, with $75.2 billion in quarterly data center revenue versus AMD’s $5.8 billion, a ~75% gross margin, and an estimated 80–90% share of the AI accelerator market. AMD is the clear number two and closing the technical gap with the MI455X and Helios rack-scale systems, but Nvidia’s scale, margins, and CUDA software ecosystem keep it dominant.
What is the AMD Instinct MI450?
The AMD Instinct MI450 is AMD’s next-generation data center AI accelerator, part of the MI400 series. The flagship MI455X offers 40 petaflops of FP4 compute, 432 GB of HBM4 memory, and 19.6 TB/s of bandwidth. It anchors AMD’s Helios rack-scale platform (up to 3 AI exaflops per rack) and the first OpenAI and Meta deployments, with production ramping in the second half of 2026.
What is AMD’s guidance for Q2 2026?
AMD guided to Q2 2026 revenue of $11.2 billion, plus or minus $300 million – about 46% year-over-year growth – with non-GAAP gross margin near 56%. It also expects server CPU revenue to grow more than 70% year over year, and continues to see the total server CPU market exceeding $120 billion by 2030.
What are the biggest risks to the AMD bull case?
Customer concentration (OpenAI and Meta dominating Instinct demand), execution risk on the aggressive H2 2026 Helios/MI450 ramp, Nvidia’s superior margins and CUDA moat, and tight HBM4 memory supply amid the broader memory shortage. Most of the promised deal revenue is milestone-based and not yet shipped or paid.
Why did AMD give OpenAI and Meta stock warrants?
The warrants align AMD’s biggest customers as prospective shareholders and only vest as GPUs ship and the stock rises. The structure is largely self-funding: maximum dilution – up to roughly 320 million shares combined – only triggers in the scenario where AMD has already captured enormous demand and its share price has climbed sharply.
Related Coverage
- Nvidia Earnings: $81.6B Quarter, 92% DC Surge [2026]
- AMD AM5 Socket Locked to 2029 as Intel Resets [2026]
- Nvidia RTX Spark: 1-Petaflop Chip Hits Intel, AMD [2026]
- Memory Chip Shortage 2026: HBM Takes 23% of DRAM Wafers
- DDR5 RAM Prices Up 110% as AI Devours Memory [2026]
- More hardware & AI chip coverage
Reporting current as of June 04, 2026. Financial figures reflect AMD’s Q1 2026 results (reported May 5, 2026) and Nvidia’s Q1 fiscal 2027 results (reported May 20, 2026). Stock performance and market data are point-in-time and subject to change.
