AMD: Accelerating the Future of High-Performance Computing and AI

Investment Thesis

Advanced Micro Devices (NASDAQ: AMD）is a leader in high-performance computing and AI, leveraging its EPYC CPUs, Instinct GPUs, and Xilinx FPGAs to capture growing demand in data centers, cloud, and edge computing. With Q1 FY2025 revenue of $7.44 billion (+36% YoY), driven by a 57% surge in Data Center sales, AMD’s competitive pricing and open-source ROCm platform position it as a strong alternative to Nvidia, with potential to gradually increase its AI GPU market share.

Despite risks from Nvidia’s dominance and cyclical semiconductor demand, AMD’s partnerships with AWS, Microsoft, and Google, along with its diversified portfolio, ensure robust growth. Based on the DCF model and taking into account the valuation premium due to favorable factors such as the accelerated growth of AI business and the relaxation of export restrictions, the fair value range is set at $130 to $180 per share.

Source: TradingKey

Financials — Quarterly Highlights (Q1 FY2025)

AMD’s Q2 2025 results reveals its strong position in HPC, AI, data centers, and gaming, marking significant gains despite regulatory and competitive headwinds.

· Revenue: $7.44 billion, representing a 36% YoY increase.

· Adjusted EPS: $0.96 per diluted share, reflecting improved profitability and operational leverage.

· Gross Margin: Non-GAAP gross margin was 54%, improving 2 percentage points from the prior year. GAAP gross margin stood at 50%.

· Operating Income: Non-GAAP operating income was $1.8 billion, approximately a 24% operating margin. GAAP operating income was reported at $806 million.

· Net Income: Adjusted net income totaled $1.6 billion, up 55% YoY. GAAP net income was $709 million.

· R&D Expense: Increased to approximately $1.73 billion, or around 23% of revenue, reflecting ongoing investments in AI, HPC, and advanced technologies.

· Share Repurchases: AMD has a total share repurchase authorization of approximately $10 billion (including $6 billion newly authorized in 2025).

Company Overview

AMD, founded in 1969 and headquartered in Santa Clara, California, designs high-performance processors that power many of today’s data-driven applications. Its products can be grouped into three main categories:

CPUs: Ryzen processors for consumers and businesses (desktops and laptops). EPYC processors serve data centers, providing energy-efficient and powerful computing for cloud providers and enterprises.

GPUs: Radeon GPUs target gaming and professional graphics needs, while Instinct GPUs focus on AI and high-performance computing, competing with Nvidia’s AI accelerators.

Adaptive and Embedded Solutions: Through its acquisition of Xilinx, AMD offers adaptive FPGAs used in edge computing, automotive, and industrial applications, beyond traditional CPUs and GPUs.

AMD uses a fabless model to speed innovation and reduce costs. The company works closely with major cloud providers like Amazon, Microsoft, and Google, as well as gaming console makers for broad market reach. Its open-source ROCm software platform supports AI and HPC and is growing as an alternative to NVIDIA’s CUDA.

Competitor Analysis

AMD faces strong competition in AI, graphics, and data center chips from Nvidia and Intel. Nvidia dominates the discrete GPU market with a 88% share as 2024, while AMD holds about 12%, a gap widened by Nvidia’s early launch of its RTX 50 series compared to AMD’s later RDNA 4 release. Nvidia’s biggest strength is its complete AI ecosystem built around the widely adopted CUDA platform, which creates developer loyalty and smooth integration with machine learning frameworks. AMD’s open-source ROCm platform is less mature and sometimes seen as more challenging to use, but it attracts customers who value flexibility, customization, and lower costs.

Source: PCViewed.com

Intel remains a distant third in discrete GPUs and servers, with limited market impact despite ongoing investments. However, AMD is steadily gaining ground in CPUs, increasing its market share in desktops, servers, and mobile devices.

Source: Tom's Hardware

AMD distinguishes itself with competitive pricing, strong price-to-performance, energy efficiency, and a broad product portfolio combining CPUs, GPUs, and adaptive FPGAs. Despite challenges in overtaking Nvidia’s AI dominance, AMD is viewed as the main alternative for cloud providers and enterprises that want to diversify or prefer open, customizable hardware solutions.

Revenue Breakdown

Source: AMD, TradingKey

AMD’s revenue in Q1 2025 was driven by four main segments, with the Data Center segment leading growth thanks to robust AI and cloud demand.

The Data Center segment, encompassing EPYC CPUs and Instinct AI GPUs, generated approximately $3.674 billion (about 49% of total revenue), marking a strong 57% YoY increase due to expanded adoption by cloud and enterprise customers. This high-margin growth highlights AMD’s strategic pivot toward AI and cloud computing, positioning the company to capture a larger share of the AI chip market.

The Client & Gaming combined segment reported revenue of roughly $2.9 billion (39% of total revenue), growing 28% YoY. Within this, Client revenue alone (laptops and desktops, primarily Ryzen CPUs) accounted for approximately $2.3 billion (31% of revenue), surging 68% YoY driven by strong demand for Zen 5-based processors and higher average selling prices. The Gaming revenue, largely Radeon GPUs and semi-custom console chips, was about $647 million (8.7% of total revenue), declining 30% YoY due to the weakening console cycle and tariff impacts.

The Embedded segment, which includes Xilinx adaptive FPGAs for automotive and industrial markets, contributed $823 million (11% of revenue), slightly down by 3% YoY amid broader economic softness in industrial sectors.

Growth Potential

AMD is well placed to grow in fast-expanding tech areas like AI, high-performance computing, and edge intelligence. A key factor in AMD’s growth potential is its expanding AI market share. Currently holding roughly 12% of the AI GPU market, AMD has the opportunity to increase this to approach 15%, driven by cost advantages and partnerships with leading cloud providers like Microsoft Azure and AWS. Its latest AI accelerators, the Instinct MI350 released in 2025 and the upcoming MI400 in 2026, use advanced chiplet designs and high-speed memory to compete with Nvidia’s lead in AI training and inference. As cloud providers and businesses look for cheaper and more available options beyond Nvidia, AMD is gaining ground by offering strong performance at better prices.

Beyond data centers, AMD’s acquisition of Xilinx has opened new growth opportunities in adaptive computing. The integration of Xilinx’s FPGAs expands AMD’s reach into edge AI, automotive, industrial, and IoT applications, diversifying its revenue base and creating synergies with its CPU and GPU offerings.

Valuation

Discounted cash flow (DCF) models incorporating current segment trends and AI-driven growth suggest a fair value range of $100 to $126 per share. Given recent positive developments: the easing of export restrictions allowing Instinct MI308 shipments to China, a modest premium is justified, supporting a target price range of $130 to $180.

Risks

AMD faces multiple key risks:

· Competitive risks: Nvidia’s AI GPU dominance and strong CUDA ecosystem make it hard for AMD to scale quickly, especially among developers and enterprise AI users.

· Software ecosystem: ROCm is less mature than CUDA, with bugs and integration issues that slow customer adoption.

· Macro and cyclical risks: Semiconductor demand is highly volatile, affected by swings in PCs, gaming, and cloud spending.

· Supply chain risks: AMD’s fabless model adds agility but exposes the company to global foundry constraints and geopolitical disruptions.

· Valuation risk: Shares have recently traded at premium multiples, with execution risk if adoption rates or margin expansion fall short of bullish assumptions.