Intel CPUs in Short Supply, Company-Wide Price Hikes Restore Pricing Power
Claire Weston
Intel has raised list prices on both Xeon server and Core Ultra laptop chips — its first supply-driven price hike in years. The CFO says last quarter's server CPU revenue grew 20%-25% year-on-year, driven by higher average selling prices, not volume — a sign the company has reclaimed pricing leverage.
What exactly is Intel raising prices on?
Intel lifted official prices on both Xeon server chips and Core Ultra laptop chips ahead of Q3, triggered by production capacity falling behind orders.
Wedbush analyst Matt Bryson notes the key question is not *whether* Intel can raise prices, but where the increase lands — OEM procurement prices moving up in tandem, or mainly retail and distribution channels.
This means → Server chips carry a larger share of Intel's revenue, so this round of increases will boost profits primarily on the enterprise side, not through retail prices consumers feel directly.
Is revenue growing from higher prices or higher volumes?
CFO Dave Zinsner disclosed that server CPU revenue grew roughly 20%-25% year-on-year last quarter, driven by rising average selling prices — not shipment growth.
The deeper signal: Intel also saw price recovery on a per-core, like-for-like basis — a metric that had been declining for years. In plain terms = chips aren't just selling for more because they're bigger; the same spec is fetching a higher unit price.
Zinsner's own words: "You could probably make a CPU today and sell it." He added that demand can sustain growth for the next three years; the constraint is supply, not willingness to buy.
How is agentic AI changing the role of CPUs?
The CPU-to-GPU ratio inside AI systems is shifting dramatically: training-era setups typically ran 1 CPU to 8 GPUs; in agentic AI — a new paradigm where AI autonomously executes multi-step tasks — the ratio is approaching 1:1, with some cases hitting 4 CPUs to 1 GPU.
This reflects how agentic workloads operate: frequent data retrieval, code execution, and rule validation — tasks CPUs handle better than GPUs.
This means → Widespread adoption of agentic AI would structurally expand the CPU demand base, shifting the demand curve upward rather than relying on cyclical restocking alone.
When will the capacity bottleneck ease?
Intel has pulled forward the yield-qualification timeline for its next-generation 18A process node — an advanced chip manufacturing technology — by at least one quarter, which should accelerate chip output.
The company has also set an internal long-term financial target called the "Rule of 45": revenue growth rate plus operating margin exceeding 45%, described by the CFO as a multi-year goal.
In plain terms = capacity is chasing demand, but has not caught up — the supply-short dynamic is unlikely to change in the near term.
How does Wall Street view the story?
Of 46 covering analysts, 12 recommend buying, 32 are neutral, and 2 rate a strong sell; the average price target is $102.87, below the $112 share price at the time of reporting.
This means → Most analysts believe the current share price has already priced in the upside from the price hikes, and consensus leans cautious.
The single most important test ahead: once new capacity comes online, can Intel hold its pricing power? If supply catches up with demand, the pricing narrative weakens.
Content is for reference only, not financial advice.