Gotrade News - For years, playing the AI-chip boom really came down to one thing: owning Nvidia. That just got more complicated. Advanced Micro Devices (AMD) rolled out its Helios rack-scale system this week, aimed squarely at the slice of the market Nvidia has long owned. Helios runs on Instinct MI455X GPUs, EPYC Venice CPUs, Pensando networking, and ROCm software. Then came the part that got Wall Street's attention. According to Barchart, Microsoft agreed to run Helios on Azure for its frontier AI inference workloads. Traders reacted fast. AMD stock popped more than 8% on July 21, the day after the news broke, and it's now up 257% over the past year. Here's why it matters for anyone holding US stocks: the AI build-out may finally have more than one obvious winner.
Key Takeaways
Microsoft will run AMD's Helios platform on Azure, lining up next to Meta, Oracle, and OpenAI as an Instinct customer.
The news pushed AMD up more than 8%, and the stock has now gained 257% in a year to around $551.66.
Qualcomm crashed the party too, signing three hyperscalers and chasing $15 billion in AI revenue by fiscal 2029.
Nvidia still leads, but it heads into its August 26 earnings carrying sell-the-news risk after a 14% run this year.
The timing helps AMD, because the business was already heating up well before Microsoft signed on. Per Barchart, first-quarter 2026 revenue hit $10.3 billion, up 38% year over year. Non-GAAP EPS landed at $1.37, a 43% jump. And the growth is lopsided in a good way: the Data Center segment alone brought in $5.8 billion, up 57% from a year ago. That's the corner of AMD that CEO Lisa Su keeps tying back to the MI450 series and Helios, the products now working their way into hyperscaler data centers.
Microsoft Bets on AMD's Helios Rack for Azure
Why does one more customer matter this much? Because Microsoft (MSFT) runs one of the largest AI training and inference operations on the planet, and it doesn't hand that kind of workload to unproven silicon. According to Barchart, the Helios deployment targets frontier AI inference on Azure, exactly where accelerator demand is climbing fastest. Landing Microsoft, on top of Meta, Oracle, and OpenAI, tells every other hyperscaler that AMD's stack can hold up at scale.
Investors clearly bought the story. AMD now changes hands near $551.66, roughly 6% below its record high, worth about $887 billion. Analysts are split on how much room is left. The mean price target sits at $543.83, just under where the stock trades today, while the most bullish call reaches $755, about 37% higher. The consensus rating is still a Strong Buy.
AMD's Valuation Runs Hot at 80x Forward Earnings
None of this comes cheap. Per Barchart, AMD trades at roughly 80.6 times forward earnings and 23.7 times sales, which leaves almost no cushion if execution slips even a little. Here's how the three US-listed chip names stack up as the race widens.
Company
Newest AI Platform
Key Metric
Stock Reaction
AMD
Helios / Instinct MI455X
Data Center revenue +57% YoY to $5.8B
+8% on July 21
Nvidia
Existing GPU accelerators
+14% YTD, target ~$304
Down ~1.49% into earnings
Qualcomm
AI200 / AI250, Dragonfly C1000
Targeting $15B AI revenue by FY2029
Down 1.70% to $172.64
Qualcomm Enters the Data Center With Three Hyperscaler Deals
And this isn't just an AMD-versus-Nvidia story anymore. Qualcomm has muscled into the AI data center as well. Per The Motley Fool, it just signed chip deals with three hyperscalers: Microsoft, Meta, and one it won't name yet. Meta plans to build its next-generation server fleet around Qualcomm's Dragonfly C1000 data-center CPU, a tie-up both sides call multi-generational. Microsoft takes a different route, pairing Qualcomm's HBC chips with the AI200 and AI250 accelerators on Azure starting next year.
The ambition here is anything but small. The Motley Fool notes Qualcomm is gunning for at least $15 billion in AI data-center revenue by fiscal 2029, up from basically nothing a year earlier, against $44.3 billion in total revenue for fiscal 2025. Its pitch to customers is efficiency: hardware that, by its own claim, delivers four to eight times more compute per watt than today's GPU setups. The stock sits at $172.64, down 1.70%, for a $185 billion market cap, and analysts peg fair value at $228.57, about 33% higher.
Nvidia Faces Sell-the-News Risk Into August 26 Earnings
So where does all this leave the incumbent? In a tricky spot, and not only because of the new competition. Nvidia (NVDA) trades near $208.91, down about 1.49% on the day, yet still up roughly 14% this year and 6% over the past month, with a consensus target north of $304. According to Barchart, its Q2 report lands August 26. That sets up the classic sell-the-news trap: when a stock has already priced in greatness, even a strong print can drag it lower.
Chip Sector Valuations Signal Caution for Retail Buyers
One word of caution before anyone piles in. These valuations were stretched to begin with. Per Barchart, the VanEck Semiconductor ETF trades around 50 times trailing earnings after a roughly 325% run over five years, and a Renaissance Macro study found heavy inflows during weak stretches paid off only 29% of the time over the following 63 days. The broader read for US-stock investors is genuinely encouraging: more real contenders in AI chips means more ways to play the theme. Just remember that with AMD, Nvidia, and Qualcomm all priced for plenty to go right, there isn't much margin for error if the capex wave cools.
Gotrade is the trading name of Gotrade Securities Inc., which is registered with and supervised by the Labuan Financial Services Authority (LFSA). This content is for educational purposes only and does not constitute financial advice. Always do your own research (DYOR) before investing.
