Here’s something nobody had on their 2026 bingo card: Intel outperforming Nvidia. Not by a little, either. By a factor of roughly 14x.

Through the first half of 2026, Intel’s stock surged between 200% and 220% year-to-date, while AMD climbed approximately 114%. Nvidia, the company that basically became synonymous with the AI trade, managed a comparatively modest 15% gain. That’s its weakest first-half showing since 2022.

Nvidia still commands roughly 70% to 81% of the AI accelerator market by revenue. It’s not losing the war. But investors are clearly betting that the next phase of AI infrastructure won’t be a one-horse race.

CPUs are having their moment

AMD reported nearly $5.8 billion in data center revenue for Q1 2026, a 57% year-over-year increase. Intel’s data center and AI segment exceeded $5 billion in the same quarter.

AMD’s EPYC server CPUs have been a particular beneficiary. Agentic AI workloads, where autonomous AI systems handle complex multi-step tasks, lean heavily on CPU performance for orchestration and inference. GPUs remain essential for training, but the inference side of the equation is where the growth is accelerating.

Intel is riding momentum under new leadership and generating excitement around its upcoming Crescent Island chip.

Why crypto should be paying attention

A growing number of mining operations have pivoted toward AI compute hosting, repurposing their GPU farms and data center infrastructure to serve enterprise AI workloads. When chip pricing dynamics shift, as they do when AMD and Intel create more competitive pressure against Nvidia, these operations benefit from declining hardware costs.

Protocols that aggregate distributed GPU and CPU resources for AI tasks care deeply about the cost curve of the underlying hardware. More competition among chipmakers means cheaper chips, which means lower barriers to entry for node operators, which means more supply on decentralized networks.

Intel’s Crescent Island chip is particularly relevant here. If it delivers competitive performance at lower price points, it could meaningfully reduce costs for decentralized GPU networks. Many of these networks were originally built on Nvidia hardware, and diversification of compatible silicon would be a structural positive for their economics.

What this means for investors

Nvidia’s bull case remains intact on fundamentals. A 70% to 81% market share in AI accelerators is dominant by any measure. Its stock just isn’t keeping pace with the explosive moves in AMD and Intel, which were both coming off much lower bases and benefiting from narrative shifts around CPU demand.

If AMD and Intel continue gaining share in data center AI, the competitive pressure should keep hardware costs on a downward trajectory. That’s unambiguously positive for decentralized compute protocols like Render, Akash, and io.net, which depend on affordable hardware to attract node operators.

Former Bitcoin miners who pivoted to AI hosting sit at the intersection of crypto capital markets and AI infrastructure demand. Lower chip costs improve their margins, while growing enterprise demand for inference compute provides revenue stability that mining never offered.

AMD faces its own challenge in maintaining 57% year-over-year growth rates as the base effect compounds.