AMD Q1 Earnings: Data Center Revenue Surges 57% Driving Results Above Estimates, Shares Rise Over 16% After Hours

[Source: AMD Official Earnings Report]

Earnings data indicates that the primary highlight of first-quarter results was the Data Center business, which achieved revenue of $5.8 billion, a 57% year-over-year increase, representing approximately 56% of total revenue and serving as the primary driver of AMD's revenue and profit growth. FactSet data shows this figure significantly exceeded analyst estimates of $5.64 billion.

Notably, AMD's server CPU revenue surpassed that of Intel ( INTC) in terms of total Data Center revenue for the first time this quarter, with EPYC alone contributing more than 50% year-over-year growth.

AMD CEO Lisa Su explained the growth logic during the earnings call, explicitly noting that inference and intelligent agent AI are driving demand for high-performance CPUs and accelerators.

This implies that AI computing demand has shifted from centralized large model training to larger-scale distributed inference and decision-making scenarios. As AI enters an "agentic" phase requiring continuous computation and memory, cloud service providers' demand for servers featuring both high core counts and high memory bandwidth has surged, driving orders for both CPUs and GPUs.

Consequently, the company's flagship EPYC processors continue to receive substantial orders from cloud service providers. This Zen-based CPU has become AMD's cash cow, providing sustained revenue generation, while the Instinct series GPUs, targeting the AI accelerator market, are the company's fastest-growing product line. This product is not only used for AI training but is also accelerating penetration into the larger inference market.

The simultaneous volume growth of both CPU and GPU product lines has pushed Data Center segment revenue to new highs, and the high-margin nature of the Data Center business is lifting the company's overall profitability.

Furthermore, regarding future growth, Lisa Su is betting on the next-generation product matrix. The MI450 series is the mainstay of AMD's next-generation Data Center GPUs, focusing on both training and inference, while the Helios platform deeply integrates CPUs and GPUs, building a competitive advantage through architectural differentiation.

Lisa Su stated that customer engagement for the next-generation MI450 series and Helios platform continues to rise, with order expectations from major customers exceeding the company's forecasts and the pipeline for large-scale deployment projects expanding. If these orders translate into actual shipments, they can effectively alleviate market concerns regarding product gaps and competitive pressures, while providing visibility for growth in future quarters.

Notably, AMD raised its long-term outlook for the CPU market during its earnings call, significantly boosting its forecast for the Data Center CPU market's compound annual growth rate from 18% to over 35%, and projecting the market size to exceed $120 billion by 2030. Additionally, the company expects second-quarter server CPU revenue to rise by more than 70% year-over-year.

The aforementioned targets convey AMD's strong confidence, but the long time horizon also means the company will face sustained market scrutiny over the coming years.

Beyond the Data Center, total revenue for the Client and Gaming segments reached $3.605 billion, up 23% year-over-year. Within this, the Client business grew 26% and Gaming grew 11%, though a decline in semi-custom chip revenue partially offset the growth. The Embedded segment's revenue reached $873 million, up 6% year-over-year, primarily due to a recovery in several end markets.

On the profitability front, the company's first-quarter free cash flow reached a record quarterly high of $2.566 billion, more than tripling from $727 million in the same period of 2024; the free cash flow margin rose from 10% to 25%. GAAP net income reached $1.383 billion, up 95% year-over-year; operating income was $1.476 billion, up 83% year-over-year.

The data indicates that AMD is achieving simultaneous expansion in profitability while rapidly scaling its business. CFO Jean Hu stated in a statement that the strong first-quarter financial performance, characterized by accelerating revenue growth and expanding profit scale, demonstrates the company's growth momentum and the advantages of its operating model.

Behind the strong earnings, Wall Street remains concerned about AMD's capacity. HSBC analyst Frank Lee downgraded AMD from 'Buy' to 'Hold' before the earnings release, arguing that AMD's ability to deliver server CPU revenue upside in 2026 will be constrained by TSMC's tight advanced node capacity.

Market estimates suggest that if TSMC's CoWoS capacity expansion falls short of expectations, AMD's AI chip shipments in 2026 could be 15%-20% lower than potential demand.

In response, Lisa Su explicitly stated that the company is coordinating with partners to substantially increase foundry and backend assembly and test capacity, a move that to some extent addresses market concerns regarding the pace of TSMC's capacity expansion. However, there is a time lag between such statements and the actual realization of capacity, and the expansion pace of TSMC's advanced nodes is not entirely under AMD's unilateral control.

Meanwhile, Susquehanna analyst Christopher Rolland maintains an optimistic outlook, having raised AMD's price target to $375 ahead of the earnings report. He is bullish on strong demand for EPYC chips and growth in AI products, noting that the better-than-expected results and second-quarter guidance provide fresh ammunition for the bull case.

AMD and Meta ( META) signed an agreement on February 24, 2026, to deploy 6 GW of Instinct GPUs, with the first 1 GW utilizing a custom version of the MI450 GPU, and will concurrently become one of the first customers for the sixth-generation EPYC processors, Venice and Verano.

The company's strategic planning for its next-generation products, the MI450 series GPUs and the sixth-generation EPYC Venice processors, is paving the way for rapid growth in the second half of 2026.

Furthermore, Intel's next-generation Xeon processor, Diamond Rapids—originally scheduled for release in the second half of 2026—has been delayed until mid-2027. This means that during the critical window in the second half of 2026, AMD's next-generation EPYC Venice processors will face almost no direct competition from same-generation rivals.

Despite the broad earnings beat, debates over AMD's valuation are set to continue. Capacity constraints in advanced nodes such as TSMC's 3nm process remain an industry-wide challenge, while the long-cycle nature of AI infrastructure investment provides strong demand support for AMD.

Lisa Su's statement that 'customer orders have exceeded initial expectations' suggests that, at least on the demand side, the market is offering a more positive assessment of AMD's competitive position. Turning that robust demand into actual shipments and revenue will be the question AMD must continue to answer in the coming quarters.

This content was translated using AI and reviewed for clarity. It is for informational purposes only.