Micron and Intel race higher in the AI surge, but their paths diverge on memory versus compute

Micron Technology and Intel have become unlikely twin beneficiaries of the artificial intelligence boom, with both stocks rising more than 200% year to date, according to a recent market column. The comparison, however, is less about who is up the most and more about what investors think comes next for AI infrastructure, where memory chips and compute chips play different roles and face different bottlenecks.

Micron is positioned in the market as a pure-play on memory, a category that has been in unusually high demand as AI accelerators scale. In the analysis, Micron’s outlook is tied to the pace of revenue growth it is posting now, while also benefiting from the idea that AI systems require large amounts of high-bandwidth memory to keep up with training and inference workloads.

Intel’s bullish case is framed around its attempt to turn multiple parts of its portfolio into an AI advantage, including CPUs at the center of data-center workloads, AI-focused products, and its foundry services. The column also points to Intel offering application-specific integrated circuits, or ASICs, which it characterizes as an alternative approach to general-purpose graphics processing units (GPUs) for certain AI tasks.

One of the more specific claims in the market discussion is about how “agentic AI” may change chip mix requirements. Agentic AI refers to systems that can take multi-step actions rather than just generating a single output. The column argues that these more complex systems could increase CPU usage relative to GPUs in a way that benefits Intel, suggesting a shift in the expected CPU-to-GPU ratio from roughly one CPU per eight GPUs to something closer to one CPU per one to two GPUs, even while acknowledging that AI data centers would not stop buying NVIDIA GPUs.

The column ties that chip-mix argument back to Intel’s reported business results. It says Intel posted 7% year-over-year revenue growth across its business, while its Data Center and AI segment rose 22% year over year and accounted for more than one-third of total revenue. Those figures, as presented in the market write-up, are used to suggest that demand for AI-related compute is not only a narrative, but also showing up in segment-level growth.

For Micron, the near-term emphasis is on “explosive” revenue growth rates, again as described in the column, with the overall conclusion being that the company may offer the better setup right now because it is delivering the more immediate fundamental payoff from the AI build-out. That contrasts with the Intel case, which is portrayed as more dependent on scaling AI-related products across CPUs, foundry, and other accelerators over time.

Still, the article’s central question is also a reminder that both companies are exposed to the same broad capital-spending cycle in semiconductors. If AI data-center investment slows, or if supply and pricing conditions for memory and compute change faster than demand growth, either stock can re-rate quickly. Neither the market column nor the supporting research consulted here provides a detailed, segment-by-segment forecast for the next few quarters, so the debate ultimately rests on whether current growth rates can persist and whether Intel’s AI-related mix shift plays out as expected.