Analyst(s): Futurum Research
Publication Date: February 5, 2026
AMD closed FY 2025 with strong data center and client execution, underscored by broad EPYC adoption, Instinct GPU traction, and an expanding AI software stack. Management outlined a 2H FY 2026 inflection as MI450 and Helios rack-scale systems enter volume and server CPUs continue to outgrow the market.
What is Covered in this Article:
- AMD’s Q4 FY 2025 financial results
- Data center AI ramp and MI450/Helios
- EPYC share gains and cloud instances
- Client AI PC and gaming outlook
- Guidance and Final Thoughts
The News: Advanced Micro Devices (NASDAQ: AMD) reported record Q4 FY 2025 revenue of $10.3B, up 34% year-on-year (YoY), versus consensus of $9.7B; importantly, the revenue beat would have held even excluding the $390M in China-related MI308 sales that were not included in Q4 guidance. By segment, Data Center revenue was $5.4B (+39% YoY), Client revenue was $3.1B (+34% YoY), Gaming revenue was $843M (+50% YoY), and Embedded revenue was $950M (+2.9% YoY). Non-GAAP operating income was $2.9B (+41% YoY), with a non-GAAP operating margin of 28% versus 26% a year ago. Non-GAAP net income was $2.5B (+42% YoY) and non-GAAP diluted EPS was $1.53 as compared to $1.09 in Q4 FY 2024.
“2025 was a defining year for AMD, with record revenue and earnings driven by strong execution and broad-based demand for our high-performance and AI platforms,” said Dr. Lisa Su, AMD chair and CEO. “We are entering 2026 with strong momentum across our business, led by accelerating adoption of our high-performance EPYC and Ryzen CPUs and the rapid scaling of our data center AI franchise.”
AMD Q4 FY 2025: Record Data Center And Client Momentum
Analyst Take: AMD exits FY 2025 with a clear data center growth vector on both CPUs and GPUs, reinforced by an expanded AI software ecosystem and rack-scale systems strategy. The company’s server CPU order book strengthened into early FY 2026, while Instinct adoption broadened across top AI customers as MI350 ramps give way to MI450 and Helios in H2 FY 2026. Client results benefited from multi-gen Ryzen demand and a growing commercial mix, even as management models a softer PC TAM given component cost inflation. Gaming returns to a normalized late-cycle cadence for semi-custom in FY 2026, while Embedded stabilizes on improving demand and a strong design-win pipeline. Together, these dynamics set up a H2 FY 2026 inflection focused on rack-scale AI, continued CPU share gains, and diversified growth vectors across PC and embedded.
Data Center AI Ramp: From MI350 To MI450 And Helios
Instinct MI350 shipments accelerated in Q4 FY 2025 as adoption broadened to eight of the top ten AI companies and neoclouds expanded on-demand access to AMD infrastructure. AMD continues to expand ROCm support, including upstream integration with vLLM and domain-specific frameworks in healthcare, streamlining time-to-deploy and performance at scale. Management highlighted MI450 and the Helios rack as the next inflection, with revenue recognition at shipment to rack builders and volume weighted to Q4 FY 2026 after a Q3 start. AMD is pursuing multi-year, multi-gigawatt deployments, including a 6 gigawatt partnership with OpenAI, and is engaged with multiple hyperscalers and enterprises for MI450-based ramps. Portfolio breadth spans MI455X and Helios for superclusters, MI430X for HPC and sovereign AI, and MI440X for enterprise eight-GPU systems, addressing a wide range of training and inference workloads. These elements collectively point to a H2 FY 2026 AI acceleration as rack-scale deployments scale across cloud and enterprise.
EPYC Momentum: Share Gains, Platform Depth, And Supply Readiness
Server CPU demand remained strong, with fifth-gen EPYC Turin accounting for more than half of server revenue in Q4 FY 2025 and prior-gen Genoa still competitive on performance and TCO across workloads. Cloud momentum continued as hyperscalers launched more than 500 AMD-based instances in FY 2025, lifting total EPYC instances to nearly 1,600, while North American hyperscalers expanded deployments. Enterprise traction improved on performance leadership, broader platform availability, and software enablement, with more than 3,000 EPYC-based OEM solutions in the market and large enterprises deploying on-prem more than doubling in FY 2025. Management expects the server CPU TAM to grow at strong double-digits in FY 2026, and cited increased CPU supply capacity and a sequential CPU revenue increase into Q1 FY 2026 despite seasonal headwinds. The next-gen Venice CPU launches later in FY 2026, intended to extend performance, efficiency, and TCO leadership across cloud and enterprise. The setup implies continued server share gains and ASP resilience through FY 2026.
Client, Gaming, And Embedded: Mixed Near-Term, Strategic Long-Term
Client posted record revenue on multi-gen Ryzen demand, with desktop setting quarterly records and commercial sell-through up more than 40% YoY; management added AI PCs via Ryzen AI 400 and the Ryzen AI Halo developer system (up to 128 GB unified memory, up to 200B-parameter models locally). While AMD expects PC TAM softness in FY 2026 due to component cost inflation, it still models AMD client growth led by enterprise and premium mix. In Gaming, Radeon RX 9000 GPUs saw healthy holiday sellout and FSR4 Redstone launched; semi-custom is expected to decline by a significant double-digit percentage in FY 2026 given the seventh year of the console cycle, with next-gen Xbox progressing toward a FY 2027 launch. Embedded improved sequentially with revenue up 3% YoY, aided by test and measurement, aerospace, and stronger channel sell-through, alongside $17B of FY 2025 design wins and over $50B since the Xilinx acquisition. Product additions spanned Versal AI Edge Gen 2, Spartan Ultrascale+ devices, EPYC Embedded 2005, Ryzen P100, and Ryzen X100 for varied edge and physical AI use cases. These segments provide diversified growth vectors that can support total company momentum through FY 2026.
Guidance and Final Thoughts
Q1 FY 2026 guidance calls for revenue of $9.5B to $10.1B (midpoint $9.8B versus consensus of $9.4B), up ~32% YoY and down ~5% QoQ, including approximately $100M of MI308 sales to China and a non-GAAP gross margin of ~55% (consensus estimate: 54.5%). China-related revenue will be guided explicitly going forward. Management noted data center (CPUs and GPUs) up sequentially, with client, gaming, and embedded showing seasonal declines; OpEx is guided to ~$3.05B, with a 13% non-GAAP tax rate and diluted shares of ~1.65B. Beyond Q1, AMD reiterated confidence in a 2H FY 2026 AI inflection as MI450 and Helios ramp in volume, while server CPUs continue to benefit from AI-driven workloads that require high-performance x86 compute alongside accelerators. Longer term, AMD sees Data Center segment revenue growing at more than 60% annually over the next three to five years and scaling AI revenue to tens of billions in FY 2027, supported by the MI400 roadmap and MI500 (CDNA6 on 2-nanometer with HBM4E in 2027). With continued investment in hardware, software, and rack-scale systems, AMD aims to drive operating leverage and earnings expansion in FY 2026.
See the full press release on AMD’s Q4 FY 2025 financial results on AMD’s website.
Declaration of generative AI and AI-assisted technologies in the writing process: This content has been generated with the support of artificial intelligence technologies. Due to the fast pace of content creation and the continuous evolution of data and information, The Futurum Group and its analysts strive to ensure the accuracy and factual integrity of the information presented. However, the opinions and interpretations expressed in this content reflect those of the individual author/analyst. The Futurum Group makes no guarantees regarding the completeness, accuracy, or reliability of any information contained herein. Readers are encouraged to verify facts independently and consult relevant sources for further clarification.
Disclosure: Futurum is a research and advisory firm that engages or has engaged in research, analysis, and advisory services with many technology companies, including those mentioned in this article. The author does not hold any equity positions with any company mentioned in this article.
Analysis and opinions expressed herein are specific to the analyst individually and data and other information that might have been provided for validation, not those of Futurum as a whole.
Other insights from Futurum:
AMD Q3 FY 2025 Earnings Highlight Broad-Based Compute Momentum
AMD OpenAI Partnership: Scale Win or Execution Risk at 6 GW?
At CES, NVIDIA Rubin and AMD “Helios” Made Memory the Future of AI
Author Information
Futurum Research
Futurum Research delivers forward-thinking insights on technology, business, and innovation. Content published under the Futurum Research byline incorporates both human and AI-generated information, always with editorial oversight and review from the expert Futurum Research team to ensure quality, accuracy, and relevance. All content, analysis, and opinion are based on sources and information deemed to be reliable at the time of publication.
The Futurum Group is not liable for any errors, omissions, biases, or inadequacies in the information contained herein or for any interpretations thereof. The reader is solely responsible for any decisions made or actions taken based on the information presented in this publication.
