AMD (AMD.US) Q3 revenue hits record $13.4 billion, Lisa Su forecasts 66% growth for Q4

According to Woofun AI, on November 3rd local time in the United States, AMD (AMD.US) officially disclosed its financial report for the third fiscal quarter ending September 26, 2026, with its core performance fully reaching new historical highs. Against the backdrop of the continuous surge in demand for artificial intelligence infrastructure, the company's revenue for the season reached $13.377 billion, representing a substantial year-over-year increase of 45% and a quarter-over-quarter rise of 16%, significantly surpassing Wall Street's previous estimate of $13.09 billion. CEO Dr. Lisa Su emphasized in the earnings statement that data center business revenue nearly doubled year-over-year, becoming the key engine driving the overall performance breakthrough, marking that AMD has fully entered a new era of growth centered on data centers.

The profit side also performed strongly, with net income under GAAP reaching $2.428 billion, a year-over-year surge of 95%. Non-GAAP net income reached $3.471 billion, up 77% year-over-year. Non-GAAP diluted earnings per share were $2.09, higher both than last year's $1.20 and the market expectation of $1.95. Boosted by this outperformance, AMD's stock price surged over 9% in after-hours trading, with market cap fluctuations reflecting strong investor confidence in the company’s short-term profitability.

It is noteworthy that despite revenue growth, the company did not sacrifice profit margins. Instead, it achieved simultaneous improvements in profitability through optimized product mix—an especially rare feat with intensifying competition in the hardware industry.

The simultaneous improvement in gross margin and operating profit sheds light on AMD's internal operational efficiency gains. This quarter, GAAP gross profit was $7.210 billion with a gross margin of 54%, up 2 percentage points year-over-year and flat quarter-over-quarter. Non-GAAP gross profit was $7.483 billion with a gross margin of 56%, also up 2 percentage points year-over-year. The improvements in both measures demonstrate that as revenue scales up, the company’s overall product mix profitability has also increased—rising data center revenue share is a key backdrop to this change.

On operating expenses, GAAP operating expenses for the quarter were $4.378 billion, up 25% year-over-year; non-GAAP operating expenses were $3.560 billion, up 29% year-over-year. Notably, GAAP R&D expenses were $2.803 billion, higher than last year’s $2.139 billion. The expansion in R&D indicates that AMD continues to increase investment in core technologies and products. As AI accelerators, high-performance server processors, and supporting software continue to iterate, the company needs to keep investing in R&D to maintain product competitiveness and meet the growing demand from the data center market.

In terms of operating profit, this quarter’s GAAP operating profit was $2.832 billion, up 123% year-over-year, with an operating margin rising from last year’s 14% to 21%. Non-GAAP operating profit was $3.923 billion, up 75% year-over-year, with an operating margin of 29%, up from 24% year-over-year. GAAP operating profit growth outpaced revenue growth, and the operating margin increased by 7 percentage points, indicating that this quarter's revenue increase not only expanded profit size but also led to higher operating yields for the company.

Woofun AI’s compiled data shows that explosive growth in the data center business is this earnings report's highlight, with the segment’s revenue reaching $8.546 billion, up 97% year-over-year and 27% quarter-over-quarter, raising its share of total company revenue to 64%, up significantly from 47% a year ago. It already represents AMD’s largest source of both revenue and profit, with operating profit of $2.882 billion, up 168% year-over-year. The main engines of this segment’s growth were Instinct series AI accelerators and EPYC server processors.

In its official materials, AMD disclosed that the Helios server cabinet solution—powered by AMD Instinct MI450 GPUs and 6th-generation AMD EPYC processors—has already been deployed in the AI infrastructure of customers including Anthropic, Meta (META.US), Microsoft (MSFT.US), OpenAI, and Oracle (ORCL.US) Cloud Infrastructure. As generative AI applications evolve toward Agentic AI, model inference, task scheduling, and complex computing pose higher requirements for data center computing power. GPUs handle AI training and inference, while server CPUs are responsible for system scheduling and general-purpose computing. The combined demand for these two product categories is driving growth in AMD’s data center business.

By comparison, the combined client and gaming segments had revenue of $3.722 billion this quarter, down 8% year-over-year. Within that, client business revenue was $2.949 billion, up 7% year-over-year but below market expectations—mainly benefiting from demand for laptops equipped with AMD Ryzen mobile processors as well as AMD’s commercial and high-end workstation product strategy. AMD also launched the Ryzen AI Max PRO 400 series processors for local AI computing and advanced the Ryzen AI Halo platform layout. Gaming segment revenue was $773 million, down 40% year-over-year, affected by changes in demand for semi-custom chips used in game consoles.

The embedded segment posted revenue of $1.109 billion this quarter, up 29% year-over-year and 14% quarter-over-quarter. Operating profit was $453 million, up 60% year-over-year. With gradual improvement in demand from aerospace, defense, automotive, communications, and industrial sectors, the embedded business continues to recover following previous inventory corrections. This quarter, AMD also unveiled the AMD Ross AI agent tools for embedded systems development, and began providing Versal AI Core samples for long-duration aerospace missions.

As for financial condition, as of September 26, 2026, AMD held $15.776 billion in cash, cash equivalents, and short-term investments, with total assets at $94.812 billion and shareholders' equity of $68.389 billion. This quarter, operating cash flow was $1.202 billion and free cash flow was $554 million, down 64% year-over-year. Cumulative free cash flow for the first three quarters of 2026 was $4.678 billion. Capital expenditure (CapEx) for the quarter was $648 million, higher than $258 million in the same period last year; cumulative CapEx for the first three quarters of 2026 reached $1.845 billion, compared to $752 million for the same period in 2025.

With rising demand for AI chips and server processors, the company continues to ramp up capital investment to support business expansion and product delivery. Advanced process foundry, high bandwidth memory (HBM), and advanced packaging are important links in the supply chain for AI accelerators and server processors. In the earnings materials, AMD said it is deepening cooperation with key customers and supply chain partners to support product supply for Instinct GPUs and EPYC processors. In addition to expanding its business footprint, AMD is supplementing technological capabilities through acquisitions. This quarter, the company disclosed it had reached an agreement to acquire World Labs, further strengthening its model research and AI infrastructure strategy;

and at the same time reached an acquisition agreement with Taalas to introduce customized inference chip technologies, expanding its technical capabilities in AI inference. Both deals are currently in disclosure phase and have not yet been completed.

Looking ahead, AMD expects fourth fiscal quarter revenue of about $17 billion in fiscal 2026, with a margin of error of 3%. According to the guidance midpoint, Q4 revenue will increase about 66% year-over-year and about 27% quarter-over-quarter.

This forecast is significantly higher than the market’s previous average expectation of around $16.2 to $16.3 billion, reflecting the company’s prediction that data center business will continue its strong growth momentum. In terms of gross margin, AMD expects non-GAAP gross margin of about 55% in Q4, down 1 percentage point from Q3’s 56%. Actual gross margin will still be influenced by product shipment structure, business mix, and supply chain costs. As one media commented, the short-term slight narrowing of gross margin is mainly due to high supply chain costs during mass production ramp-up of next-generation full-cabinet products, but this fits the regular profitability curve for large-scale hardware delivery and does not shake market confidence in AMD’s profitability.

CNBC pointed out that AMD's strong Q4 guidance has injected a shot of confidence into capital markets recently jittered by concerns over an “AI infrastructure investment bubble.” The report analyzes that although AMD is still behind Nvidia (NVDA.US) in the AI accelerator field, it is continuously expanding its market share through Helios server cabinet designs and technological barriers on server CPUs (EPYC series), successfully establishing itself as an indispensable “second source” in the AI computing supply chain for major technology companies.

Looking to Q4, the market's focus will be on the delivery progress and capacity coordination of the Helios server cabinet system, Instinct MI450 series GPUs, and the next-generation EPYC processors. Whether AMD can steadily convert the continued growth in AI computing demand into stable product shipments and revenue, despite advanced process and storage chip capacity constraints upstream, will be crucial for its performance over the next several quarters. As AI infrastructure construction deepens, supply chain stability and cost control will be core variables for AMD to maintain high growth. Any bottleneck in the supply chain could impact further expansion of its market share.