Semiconductors are responsible for nearly half of the S&P 500’s profit growth in Q2 2026. One sector, roughly a dozen companies, is doing the heavy lifting for an index that’s supposed to represent the 500 largest public companies in America.

The stat comes via a16z, the venture capital firm that straddles both traditional tech and crypto markets. According to their analysis, semiconductor contributions to the index’s earnings growth jumped 17% from Q1, with chip stocks now accounting for approximately 44-48% of overall EPS growth. Chip earnings surged 133% year-over-year in Q2, a number that makes the rest of the index look like it’s standing still.

The concentration problem nobody wants to talk about

Semiconductor stocks now represent 19.7% of the entire S&P 500 by index weight as of June 30, 2026. For context, that figure was around 5% in mid-2020. In six years, the chip sector has roughly quadrupled its share of the most-watched equity index on the planet.

The Philadelphia Semiconductor Index, known as the SOX, has been trading 65% above its 200-day moving average. The last time a major sector index stretched that far above its long-term trend line was right before the dot-com crash in 2000.

Why crypto investors should care about chip stocks

The same investor appetite that’s pushing Nvidia and its peers to record valuations is the force behind the surge in AI-focused crypto tokens, decentralized compute networks, and GPU marketplace protocols. When chip stocks sneeze, AI tokens catch a cold. Consider the mechanics. Institutional investors rotating capital into semiconductor equities are, by extension, validating the thesis that AI infrastructure is the defining investment of this decade. That validation flows downstream into crypto projects promising decentralized versions of the same infrastructure. Render, Akash, and similar protocols have seen their token valuations track semiconductor sentiment with increasing fidelity.

Dot-com parallels and what to watch

The comparison to the dot-com bubble deserves some nuance. In 2000, the concentration was driven by companies with minimal revenue and speculative business models. Today’s semiconductor giants are printing record profits. The 133% year-over-year earnings growth is built on actual demand from hyperscalers spending billions on AI training clusters.

When nearly half of a broad market index’s growth depends on one sector, any disruption to that sector, whether from supply chain issues, geopolitical tensions around Taiwan, or simply a deceleration in AI spending, becomes a systemic risk rather than a sector-specific one.