Advanced Micro Devices vs. Navitas Semiconductor: Here's What The Quarterly Revenue Trends of These Artificial Intelligence Companies Reveal to Investors

AMD’s relative winner status is less about headline growth and more about quality of growth: it is converting AI demand into scale economics, which should support operating leverage even if top-line growth slows from peak rates. The underappreciated risk is that the next leg is no longer purely GPU-driven; if inference shifts more workload toward CPUs and platform attach, AMD can broaden wallet share, but that also makes execution on software, ecosystem, and supply allocation the gating factor rather than raw chip demand. Near term, the stock is likely to trade on whether the current plateau is digestion or the first sign that AI demand is normalizing faster than expected.

NVTS looks like a classic “story reset” where the market is still pricing the remnants of the old business while the new addressable market has not yet demonstrated enough scale to offset dilution risk. The combination of customer-basis churn, litigation overhang, and a large equity overhang raises the probability that any revenue inflection is monetized by the balance sheet first, not the equity holder. Second-order, if power-semiconductor buyers perceive patent risk or financing stress, design wins could migrate to larger, better-capitalized rivals, extending the revenue air pocket beyond one quarter.

The contrarian view is that AMD may be less cheap than it looks because the market already rewards visible AI exposure, while NVTS may be over-penalized if AI-focused power conversion actually reaccelerates and the China exit was the right strategic prune. Still, the asymmetry favors waiting for proof on NVTS rather than anticipating it: dilution plus litigation typically compress multiple before they help earnings. Over 1-3 months, the key falsifier is a cleaner NVTS revenue rebuild without share count expansion; over 6-18 months, the key AMD falsifier is any sign that packaging spending and inference demand fail to translate into sustained margin or share gains.