Intel CEO Lip-Bu Tan has put forward a case that deserves attention: the balance between GPUs and CPUs in AI systems is shifting. Where the ratio once sat at roughly 8:1 in favor of graphics processors, Tan argues it is moving toward parity — and in some inference workloads, where already-trained models are executed, CPUs could even overtake GPUs. It is a thesis that reframes how Intel should be judged in the AI era. It is not, however, proof that a turnaround has taken hold.
That distinction — between a technical argument and an economic one — runs through the entire investment case. Intel published no sales data to back the claim. And a growing market is not the same as growing Intel profit.
Forecasts Are Not Order Books
The projections attached to the CPU narrative are ambitious. Intel’s finance chief expects industry-wide server CPU unit growth of more than 30% in both 2026 and 2027. Tan goes further, floating the possibility of the server CPU market expanding as much as fivefold.
Those are forecasts, not orders. Supply remains part of the same equation: management has flagged that shortages could persist until 2028. Strong demand and constrained delivery can coexist, and an optimistic market outlook substitutes for neither available product nor successful shipment.
What the Market-Share Numbers Actually Say
Mercury Research data published today illustrates how tricky the read-through is. Intel’s desktop CPU share climbed to 70.1% in the third quarter of 2026, up from 65.1% in the prior quarter.
That looks like regained strength, but the cause is unclear. According to media reports, Intel may have shifted shipments toward the desktop market amid weak laptop demand — meaning the higher share is not necessarily a sign of broad-based demand recovery.
At the same time, AMD reached a record 35.5% share of server CPUs, per Mercury Research, leaving Intel at 64.5%. The server processor market is expanding, but not without fierce competition over how it is divided. Even if Tan is right about the broader technology trend, that does not automatically translate into reclaimed ground. The CPU can become more important while a rival benefits too.
Should investors sell immediately? Or is it worth buying Intel?
The Numbers Behind the Narrative
The operating business already offers clues about the weight of data centers. Intel’s Data Center and AI segment generated $7.318 billion in revenue in the third quarter of 2026, up 78% year over year, with a segment operating margin of 42%. Intel Foundry, by contrast, posted a loss of $1.85 billion.
Growth opportunity and financial drag sit side by side. The AI story must not obscure either the weak PC business — Intel expects PC market softness to last at least through the first half of 2027 — or the costs of other divisions. For the fourth quarter of 2026, Intel guided to revenue of $15–16 billion and adjusted earnings per share of $0.25.
A Fresh Buy Call, and What It Rests On
A new buy recommendation landed today, with an analyst arguing that the recent price correction has improved the risk-reward balance. The case combines the operational recovery with potential tailwinds from the Nvidia partnership, which is meant to open access to accelerated-computing ecosystems through custom data-center processors and PC chips. Until revenue from that collaboration materializes, this leg of the thesis remains an expectation rather than a realized contribution.
Higher server prices and rising volumes underpin the recovery argument, and the data-center segment’s third-quarter performance gives it a reported foundation. The analyst also acknowledges the offsets: weak PC demand, supply bottlenecks, and execution risk in manufacturing.
A Fund Position, Carefully Read
Soros Fund Management raised its Intel stake substantially in the third quarter of 2026, according to Reuters, with the disclosed position valued at $340.3 million at quarter-end. The build-up fits the constructive view on recovery prospects, but it is no substitute for a reason to buy today: the filing discloses no investment rationale and reflects a position as of September 30. The lagged disclosure permits no firm conclusion about the fund’s current holding, nor any inference about a reaction to later corporate news.
The Yardstick That Matters
The return of the CPU is a serious thesis, but no blank check for the stock. The measure should not be the boldest growth forecast. It should be the visible link between rising server demand, Intel’s competitive position, and an improving earnings picture. Only there does a plausible technology story become a durable corporate trajectory.
Ad
Intel Stock: Buy or Sell?! New Intel Analysis from November 15 delivers the answer:
The latest Intel figures speak for themselves: Urgent action needed for Intel investors. Is it worth buying or should you sell? Find out what to do now in the current free analysis from November 15.
Intel: Buy or sell? Read more here...
