AMD reports second-quarter results after the close today, with the call at 5:00 p.m. ET. The stock is up roughly 125% year to date. That single fact reframes the entire event: this is not a report that has to be good, it is a report that has to be better than a market that has already paid for good.
The bar
| Metric | Street / Company |
|---|---|
| Revenue, consensus | ~$11.3B |
| Company guidance | $11.2B ±$300M |
| Implied YoY growth at midpoint | +46% |
| Implied sequential growth | +9% |
| Adjusted EPS, consensus | $1.61 |
| Stock, YTD | +125% |
| Q1 2026 data center revenue | $5.8B |
The revenue number is nearly settled — the company guided $11.2 billion and the Street landed at $11.3 billion. Beating by a hundred million is table stakes and will not move anything by itself.
What actually moves it
The MI400 ramp. AMD has committed to an annual cadence for AI accelerators: MI325X in late 2024, the MI350 series in 2025, MI400 next. At the July Advancing AI event the company introduced the MI400 series and Helios, a rack-scale system pairing Instinct GPUs with EPYC CPUs. What the market needs tonight is not another product slide — it is units, timing, and customer names.
Helios in Azure. Microsoft has committed to integrating Helios into Azure starting in the second half of 2026. That is the single most important external validation AMD has in AI, because it moves the story from "a second source exists" to "a hyperscaler has budgeted for the second source." Any color on volume there is the highest-value sentence in the call.
Second-half guidance. With 46% growth already in the run rate, the second-half guide is where the multiple gets set. A guide that implies acceleration justifies the 125%. A guide that implies the same run rate does not.
Supply. AMD is competing for the same advanced packaging and memory capacity as everyone else in AI, in a year where memory is in genuine shortage. If the constraint is supply rather than demand, that is a very different — and in some ways better — problem, but it caps the upside in the near term.
The part I'd be careful about
A 125% run into a print is its own risk factor. The distribution of outcomes on a stock like this is not symmetric: a strong quarter gets a shrug because it was expected, and a soft guide gets punished twice — once for the miss and once for the positioning unwinding.
I'd also be careful about the Nvidia comparison, which is how this stock is traded and not how it should be underwritten. AMD does not need to beat Nvidia. It needs hyperscalers to want a credible second supplier badly enough to build software for it. Those are different bets and the second one is far more achievable.
My take
I don't buy semiconductors the afternoon of a print after a double in seven months. The trade is fine; the risk-reward is not.
What I'm listening for is one specific thing: whether Helios revenue gets quantified or stays qualitative. "Ramping in the second half" is a phrase AMD has used before about accelerators. A dollar figure, a unit number, or a named second hyperscaler is what would make me treat this as a step-change rather than a strong cyclical quarter.
If the guide is strong and the stock sells off on positioning, that is the setup I'd actually want. Good news into a crowded trade often creates a better entry a week later than the entry available tonight.
Bottom line: revenue is nearly pre-agreed. The MI400 and Helios detail is the whole event, and a 125% YTD run means the burden of proof sits with the company.
This is analysis, not investment advice.