AMD’s AI Boom Hits a Wall With Margin Guidance

According to CNBC, AMD reported fiscal third-quarter revenue that jumped 36% year-over-year, beating Wall Street expectations with net income climbing to $1.24 billion from $771 million a year earlier. The company expects about $9.6 billion in revenue for the fourth quarter, which would represent 25% growth and exceeds analyst consensus of $9.15 billion. However, AMD’s adjusted gross margin guidance of 54.5% for the upcoming quarter only meets StreetAccount’s estimates rather than exceeding them. The stock slipped in extended trading despite the strong top-line performance, with shares having already gained 107% this year through Tuesday’s close. AMD’s data center business generated $4.34 billion in revenue, up 22% and beating expectations, while client revenue reached $2.75 billion, up 46%.

The AI gold rush meets reality

Here’s the thing about AMD right now: everyone wants them to be the Nvidia killer. The stock has more than doubled this year on pure AI hype. And when you’re trading on that kind of momentum, meeting expectations just isn’t good enough. The market wanted AMD to crush margin guidance, not just match it.

Basically, AMD is caught between two worlds. Their traditional businesses are doing great – data center up 22%, client revenue up 46%, gaming up a massive 181%. But the real story is whether they can actually compete with Nvidia in the AI processor space. That recent OpenAI deal where the AI giant might take a 10% stake? That’s the kind of validation investors are banking on.

The China question

Now here’s something interesting that keeps popping up quarter after quarter. AMD specifically noted that their guidance doesn’t include revenue from shipments of their Instinct MI308 chips to China. They said the same thing last quarter. So what’s really going on there?

Is this just being cautious about export controls? Or is there something bigger happening in their China business that they’re not talking about? When you’re trying to catch Nvidia, every revenue stream matters. And leaving out what could be significant China sales from guidance? That feels like there’s more to the story.

Partnerships and problems

AMD’s expanding their reach with deals like the recent Oracle partnership to help customers achieve AI scale. But then you’ve got Amazon – a key cloud customer – selling all their AMD shares as of September 30th. That’s 822,234 shares gone.

So which is it? Are big tech companies all-in on AMD’s AI future, or are some getting cold feet? Amazon building that position in Q1 only to dump it by Q3 suggests they might see something the market doesn’t. Or maybe it’s just portfolio rebalancing. But timing-wise, it’s definitely worth watching.

Where does AMD go from here?

The fundamental story hasn’t changed – AMD is executing well across multiple segments. But when your stock trades at 107% gains for the year, perfection becomes the expectation. Meeting margin guidance instead of beating it? That’s enough to send the stock down after hours.

I think the real test comes in the second half of next year when that initial 1-gigawatt rollout of Instinct chips to OpenAI begins. Until then, AMD needs to prove they can not just grow revenue, but grow profitability in the AI space. Because right now, Nvidia isn’t just winning on revenue – they’re crushing it on margins too.