Advanced Micro Devices has earned a Strong Buy rating in part because it serves as a lower-risk, diversified hedge to Nvidia exposure. Unlike Nvidia’s concentrated hyperscaler revenue, AMD derives revenue from a broader mix of gaming GPUs, client PCs and data center AI accelerators. In its latest outlook, AMD highlighted that mid-market hyperscale customers are adopting its competitively priced Instinct MI200 series, while its upcoming rack-scale solutions promise to expand total addressable market by targeting enterprise and edge deployments that Nvidia’s premium offerings often price out.

AMD’s horizontal research and development strategy—spanning CPUs, GPUs and networking—coupled with its open-source ROCm software platform is designed to erode Nvidia’s long-standing proprietary advantage. Over the past year, hyperscale customers such as Microsoft, Meta, Oracle and OpenAI have begun integrating AMD Instinct accelerators alongside Nvidia GPUs, citing lower total cost of ownership and improved software interoperability. By aligning with open standards, AMD expects to capture double-digit percentage points of market share in the AI accelerator market in 2026.

Management projects ‘monster’ revenue growth over the next several years, with data center revenue growing 36% year-over-year in the most recent quarter and full-year guidance calling for 31% growth. While AMD’s gross margins remain below Nvidia’s industry-leading levels, this gap allows AMD to price its accelerators more aggressively and win design slots in both training and inference workloads. Additionally, the company has engineered new AI chips with double the on-die memory bandwidth of prior generations, which it believes will drive further adoption in inference-focused applications and help sustain a multi-year growth runway.