The call discusses AMD's third quarter 2025 financial outcomes, focusing on non-GAAP measures. It previews upcoming analyst events, highlights forward-looking statements, and acknowledges risks impacting future performance. Key financial disclosures and strategic insights are shared.

AMD delivered record revenue and profitability, driven by strong sales in data center AI and server CPU segments, with significant growth in Epic processor and Instinct GPU sales, reflecting robust demand from hyperscalers and enterprises.

Epic platforms, optimized for enterprise workloads, see strong adoption across various sectors. With over 170 5th gen platforms in market from multiple vendors, significant wins are reported. Looking ahead, the upcoming 2nm Venice processors, currently in testing, promise enhanced performance and efficiency, backed by strong customer interest and pre-launch cloud deployments.

AMD's Instinct GPU business is experiencing significant growth, driven by the ramp-up of MI series GPU sales and deployments. The company highlights partnerships with major cloud and AI providers, including Oracle, Crusoe, DigitalOcean, and Tensor Wave, for Mi 350 series public cloud offerings. Additionally, AMD showcases progress in software with the launch of ROCm 7, which delivers enhanced performance and introduces new features for AI development. The open software strategy is gaining traction with contributions from developers, solidifying ROCm as a platform for AI development.

AMD's data center AI business is poised for significant growth with the launch of Mi 400 series accelerators and Helios rackscale solutions. The company has secured a multi-year agreement with OpenAI, establishing AMD as a core compute provider. Oracle and Cisco are also key partners, with Oracle deploying Mi 450 GPUs across its cloud infrastructure and Cisco building a large-scale AI cluster powered by Instinct Mi 350x GPUs. The US Department of Energy has selected AMD's upcoming Mi 430x GPUs and Venice CPUs for the next flagship supercomputer, solidifying AMD's leadership in powering the world's most powerful computers. This growth trajectory is expected to generate tens of billions in annual revenue by 2027.

AMD reported record third-quarter sales with significant year-over-year growth in PC processors, gaming, and embedded segments. The company highlighted strong demand for its Ryzen CPUs, Radeon graphics, and embedded solutions, alongside robust design wins and market adoption, setting a positive outlook for future growth.

Strong third-quarter financial results, with record revenue of $9.2 billion, driven by significant growth in data center and client gaming segments. Outlook for Q4 2025 includes revenue guidance of approximately $9.6 billion, reflecting continued momentum in data center and embedded segments, and strategic investments in AI opportunities.

Announcement of a Q&A session initiation with instructions on how to join the queue for questions, emphasizing one question and one follow-up per participant.

Discusses strong Q3 performance in data center AI and server sales, anticipates continued demand growth, and outlines plans for Mi 355 ramp and Mi 450 series launch in 2026.

Discusses the strategic planning and collaboration with major AI clients to ensure adequate power and supply chain readiness, highlighting efforts to model allocation and enhance visibility in critical customer engagements.

The dialogue highlights the strong market response and customer interest in the Helios system and Mi 450 following the OCP show. It discusses the crossover point for discrete versus system sales, noting increased engagement with customers who are eager to explore the system's features. Early adopters for Mi 4 are expected to focus on rack-scale solutions, with anticipation for further form factor developments.

An industry leader emphasizes the need for collective planning among the ecosystem to address power requirements and component availability issues for upcoming large-scale deployments, highlighting both data center infrastructure and power as potential gating factors.

The company is collaborating with supply chain partners to ensure sufficient capacity in silicon, memory, packaging, and components over the next two years. Despite tight conditions, significant growth in compute demand is anticipated, with efforts focused on increasing power and supply to meet the challenge.

The dialogue highlights the sustained CPU demand driven by AI workloads, predicting a positive environment into 2026. It reassures supply readiness for growth, particularly in 2026, and notes the broadening of demand among major clients, indicating a durable trend beyond seasonal fluctuations.

AMD has significantly improved Rock with R7, enhancing performance and framework support, ensuring a smooth experience for new customers. The focus is on expanding libraries and environments for emerging workloads like training, inference, and reinforcement learning. Continuous investment is pledged to maintain a competitive edge in developer support.

Discussion on the framework for improving gross margins during the transition to new data center GPU products, emphasizing top line revenue growth and margin expansion.

The dialogue touches on growth expectations for 2026 and beyond, focusing on OpenAI and large customers' role in driving tens of billions in scripts. It hints at broader customer penetration strategies, promising more detailed insights at an upcoming analyst day.

Highlights significant customer engagements including OpenAI, OCI, and Department of Energy, emphasizing the company's focus on scalable deliveries and robust supply chain preparation for multiple large-scale partnerships.

The dialogue discusses the year-over-year growth of servers and data center AI within a data center context. It concludes that both areas grew nicely, with servers exhibiting slightly better growth compared to data center AI.

A discussion unfolds around the interpretation of 'strong double digits' growth for servers and data centers, aiming to align expectations with provided guidance, while also addressing inquiries about GPU projections and Mi 350 ramp-up, emphasizing the need for clarification on previously stated forecasts.

The dialogue emphasizes the robust growth in servers and data centers, with a focus on double-digit percentage increases, suggesting satisfaction with their performance.

Announcement of long-term OpenAI collaboration, alongside showcasing Helios rack, accelerates customer interest and engagement at higher scales. NVIDIA aims to diversify its data center GPU revenue, targeting multiple customers of similar scale by 2027-2028, mitigating risks associated with customer concentration.

Discussion highlights strong server CPU demand driven by AI workloads, predicting growth in ASPs and turnover. Early engagement with next-gen CPUs indicates market readiness for advanced compute solutions, expanding the AI silicon TAM outlook.

The partnership with OpenAI is enhancing software stack robustness, with shared work on hardware, software, and systems. Feedback from large customers and AI native companies is strengthening the stack, focusing on training and inference improvements, and accelerating kernel development.

The dialogue explores the practical utilization duration of GPUs beyond standard depreciation, noting a trend towards extending the use of older models for AI compute, especially in inference. It also addresses the current status and potential impact of Mi 308 shipments, highlighting the ongoing negotiation with customers regarding demand and the expected update on opportunities in the coming months.

The dialogue revolves around preparing inventory for potential market openings, with ongoing work in process and uncertainty regarding future demand shaping inventory strategies.

AMD discusses its competitive positioning in the AI compute market, highlighting the strengths of its MiLor series and emphasizing time-to-market, total cost of ownership, and deep partnerships. The company also reflects on a unique partnership structure involving warrants, suggesting openness to creative equity-based agreements to address global processing power demands, while looking forward to leveraging its capabilities in emerging AI opportunities.