Advanced Micro Devices delivered stronger-than-expected second-quarter results as demand for data center hardware more than doubled from a year earlier, underscoring how spending on AI infrastructure is increasingly determining the company’s growth profile. AMD reported revenue of $11.536 billion for the quarter, up 50% year over year and 13% from the prior quarter, surpassing the $11.309 billion consensus estimate cited in the supplied figures.
Adjusted earnings per share reached $1.66, ahead of the $1.62 expected by the market. AMD also forecast third-quarter revenue of about $13 billion at the midpoint, compared with a $12.52 billion consensus estimate, signaling that the company expects AI-related hardware demand to remain strong through the second half of 2026.
The results place AMD’s data center division at the center of its business. Data center revenue rose 107% year over year to $6.718 billion, representing roughly 58% of total quarterly sales. The segment’s expansion far outpaced the rest of AMD’s portfolio and reflects major customers’ purchases of processors and accelerators used in cloud platforms, AI model training, and enterprise computing systems.
Data center business drives AMD’s revenue mix
The sharp rise in data center revenue shows the scale of the shift inside AMD. A year earlier, the company had a more balanced mix of exposure to PCs, gaming hardware, embedded chips, and servers. The latest quarter was dominated by infrastructure spending, with data center sales exceeding the combined revenue of the client, gaming, and embedded divisions.
Companies building AI services require far more computing capacity than conventional enterprise workloads. Training and operating large language models, image generators, recommendation engines, and other machine-learning tools relies on systems that combine advanced processors, graphics processing units, high-bandwidth memory, networking equipment, and sophisticated packaging technologies.
AMD is seeking to capture a larger share of that spending through products designed for accelerated computing and through rack-scale systems such as Helios. Rack-scale products integrate computing, networking, memory, and cooling into a larger system intended for deployment in data centers, rather than selling a single chip into a server manufacturer’s standard configuration.
That strategy can increase the value of each customer deployment, but it also creates higher initial manufacturing and integration costs. AMD reported a 56% non-GAAP gross margin and a 54% GAAP gross margin for the second quarter. The company attributed the margin pressure in part to costs associated with ramping the Helios AI rack-scale system.
Early production runs of highly integrated systems can weigh on profitability because companies must secure advanced packaging capacity, validate more components, and manage a more complex supply chain before volumes rise. AMD expects those ramp-related costs to ease as production scales, according to the information supplied.
AI competition raises the cost of participation
AMD’s results also illustrate the financial demands of competing in the AI accelerator market. Nvidia remains the dominant supplier of AI accelerator processors, with the supplied material estimating its market share at roughly 85%. That position has given Nvidia a large installed base among cloud providers and AI developers, while competitors must spend heavily on product design, software ecosystems, packaging, and customer support to win deployments.
For AMD, revenue growth alone will not settle the competitive question. The company must show that its accelerator hardware, software tools, and complete system designs can support large-scale workloads reliably enough to persuade customers to diversify beyond the market leader.
The Helios ramp is part of that effort. Building more complete systems gives AMD a route to compete for larger data center contracts, but it also exposes the company to the execution risks that come with integrating components at the rack level. The 2 percentage-point difference between AMD’s GAAP and non-GAAP gross margins also indicates that the company continues to exclude certain expenses when presenting its adjusted profitability measure.
Client and embedded divisions add support
AMD’s client segment, which includes processors for personal computers, generated $3.062 billion in revenue, up 23% from a year earlier. The gain points to healthier demand for PC processors than the gaming division recorded, although the supplied figures do not break out sales by individual product category or customer.
Embedded revenue rose 19% to $977 million. Embedded chips are used across industrial equipment, communications systems, automotive applications, and other specialized devices. The segment provides AMD with demand sources that tend to operate on longer design cycles than consumer PC or gaming products.
Gaming was the major weak point in the quarter. Revenue fell 31% year over year to $779 million. That decline should not be read solely as a measure of retail graphics card demand: AMD’s gaming segment also includes semi-custom chips used in game consoles. The figures show that gaming hardware made a much smaller contribution to AMD’s growth than data center products, without establishing how much of the decline came from consoles, discrete graphics cards, or other gaming-related products.
Limited direct read-through for crypto mining
The report has a narrower implication for cryptocurrency mining than for AI infrastructure. AI data centers often use specialized accelerators and high-end server components that are not interchangeable with the ASIC machines used to mine Bitcoin. Bitcoin miners primarily depend on application-specific integrated circuits designed for a single hashing function, rather than the general-purpose GPUs and data center accelerators central to AMD’s AI strategy.
GPU-dependent proof-of-work networks could face indirect effects if advanced chips, memory, packaging capacity, or electricity infrastructure become more expensive. Yet AMD’s quarterly results do not provide evidence that miners face an immediate hardware shortage or justify broad changes to digital asset holdings.
AMD’s third-quarter forecast instead offers a clearer message about the technology hardware market: large customers are sustaining unusually high expenditure on AI computing systems, and AMD expects to convert a growing portion of that demand into revenue even as the cost of building complete rack-scale platforms temporarily restrains margins.
Capitalize on AI and data-center growth narratives by exploring AMD-correlated crypto plays in Toobit’s markets section today.
Disclaimer: The content on this page is provided for general informational purposes only and does not represent the views or financial advice of Toobit. We make no guarantees regarding the accuracy or completeness of this information and shall not be held liable for any errors, omissions, or outcomes resulting from its use. Investing in digital assets involves risk; users should independently evaluate their financial situation and the risks involved. For further details, please consult our Terms of Service and Risk Disclosure.
