The AI chip market is not suffering one downturn but two, and the difference matters for anyone tracking semiconductor earnings this cycle. A report from NDTV Profit frames the divide starkly: memory suppliers are bleeding, while Nvidia is wobbling far less. The headline captures a real divergence in how the AI slowdown transmits through the supply chain.

Why is memory hit harder than GPUs?

Memory is a commodity business at its core. When AI server demand cools, DRAM and NAND suppliers cannot easily hold prices, because buyers can defer purchases and inventory piles up quickly. Graphics processors, by contrast, sit inside a concentrated market where Nvidia holds the dominant design position. That concentration gives it more pricing power when demand softens.

The mechanism is straightforward. Commodity parts compete on price, so oversupply cuts revenue immediately. Platform products compete on ecosystem and performance, so a demand dip slows growth rather than reversing it.

What does this mean for the chip sector?

The split suggests investors and buyers should stop treating "AI chips" as a single trade. Suppliers at different layers of the stack — memory makers versus GPU designers — face different exposure to the same demand signal. NDTV Profit's framing implies memory makers carry the heavier near-term risk, while Nvidia's wobble is comparatively shallow.

How deep each side's pain runs will depend on whether AI infrastructure spending resumes its climb or plateaus through the coming quarters.