Forget Nvidia for a Moment – AMD’s Two-Headed AI Engine Is Just Getting Started

美股半导体财报:这家芯片巨头今天给投资者送上惊喜
Published on: May 11, 2026
Author: Caroline Kong

As of May 10, 2026, Nasdaq data shows that AMD’s stock price has surged following its earnings release, with year-to-date gains exceeding 90% and the company’s market capitalization jumping to approximately $687 billion. Supporting this sharp rally is not wishful market sentiment, but clear signals repeatedly cross-verified by quarterly earnings, customer orders, and industry data: AMD is evolving from the “second choice” in the AI computing wave into an “indispensable dual-engine player.”

Earnings Beat: Data Center Becomes Core Growth Engine

AMD’s first-quarter 2026 earnings report, released in early May, surprised the market. The company achieved revenue of $10.25 billion, up 38%. Adjusted EPS came in at $1.37, also above expectations. The core driver was the Data Center segment, which posted quarterly revenue of $5.8 billion, up 57% year-over-year and 7% sequentially.

The deeper driver behind this growth is the rise of Agentic AI. AI applications are evolving from single large-model training to more complex inference and automation scenarios, dramatically increasing enterprise demand for high-performance server CPUs beyond Nvidia’s GPUs. AMD’s 5th-generation EPYC Turin and 4th-generation EPYC processors are becoming standard choices for cloud providers and enterprise customers expanding their AI clusters. Management explicitly expects data centre CPU revenue to grow more than 70% year-over-year in the second half of 2026, with continued strong growth into 2027.

Clear Product Strategy: CPU and GPU Dual Engines

Facing Nvidia’s dominant ~90% share of the AI training market, AMD is charting a differentiated yet balanced course.

On the GPU front, the upcoming Instinct MI450 series, set for mass production in the second half of 2026, leverages TSMC’s advanced 2nm process, featuring up to 432GB of HBM4 memory and approximately 19.6 TB/s of bandwidth. This product has already secured endorsements from heavyweight customers: OpenAI previously signed a multi-year collaboration framework; Meta has entered into what is likely the largest GPU procurement agreement in history – over $100 billion and a 6-gigawatt deployment plan – deeply binding custom MI450s from AMD, with the first gigawatt scheduled for delivery in the second half of 2026; Oracle has announced plans to deploy up to 50,000 MI450 chips starting in the second half of 2026 to build an AI supercluster.

On the CPU side, AMD is enjoying an equally solid and potentially even more resilient growth driver. The total addressable market (TAM) for server CPUs has been doubled by management from $60 billion to $120 billion within just six months, directly driven by the structural demand release from Agentic AI. AMD confidently states it expects to capture more than half of that market. Gartner forecasts that global semiconductor revenue will exceed $1.32 trillion in 2026, of which AI semiconductors will account for about 30%, while hyperscale AI infrastructure investment is growing more than 50%. Sinolink Securities believes AMD is transforming from a pure AI GPU vendor into a full data centre infrastructure provider encompassing “AI GPU + CPU + rack-scale solutions.”

Wall Street Turns Bullish: Analysts Raise Targets

Within days of the earnings release, major Wall Street firms quickly raised their ratings. Goldman Sachs was the most aggressive, upgrading AMD from “Hold” to “Buy” and raising its price target from $240 to $450, citing structural benefits to server CPUs from Agentic AI, plus upside from Meta’s 6GW project. BOCOM International, based on the doubled CPU TAM and the GPU ramp in the second half of the year, set a target of $468, corresponding to 39x 2027 earnings.

Valuation Has Support, But Risks Remain

After the rapid share price appreciation, valuation pressure is emerging. Based on consensus non-GAAP EPS expectations for fiscal 2026, AMD’s forward P/E is approximately 48x, with a PEG of about 1.6x – reasonable compared to pure-play AI names, but far from cheap. For a company whose AI GPU penetration is still less than 20% of Nvidia’s, the market has already priced in sustained revenue growth of over 60% annually.

Supply-side bottlenecks persist. While TSMC’s capacity support has confirmed near-term growth, AMD still faces intense competition with Nvidia, Broadcom, and others over wafer, packaging, and HBM supply. Additionally, the pace of high-end GPU deployments, whether AI infrastructure spending will maintain its high growth rate, and margin pressures from intensifying competition are all variables investors need to monitor continuously.

Bottom Line

AMD is now in a critical window where both of its growth engines are firing simultaneously: Agentic AI is driving server CPU demand to exceed expectations, while the MI450 series GPU is poised for a major deployment inflection point in the second half of 2026. For investors who can tolerate higher volatility and have an eye on the secular AI infrastructure trend, AMD is undoubtedly a core candidate worthy of close attention.

AI Growth Stocks Technology U.S. stocks