📊 Intel Q3: Data Center revenue +78 percent, Nova Lake ahead of schedule
📅 29 October 2026 • CNBC
Intel reported third quarter 2026 results that came in above analyst expectations. Revenue reached $16.51 billion, up 21 percent year over year, while adjusted EPS was $0.52 against the $0.40 that had been modeled. The strongest line item was the Data Center and AI segment, which grew 78 percent to $7.3 billion. CFO David Zinsner described it as the segment's strongest growth rate in 15 years, driven by demand for server and accelerator capacity. The segment alone now contributes roughly 44 percent of total company revenue, computed from the two reported figures. Guidance for the fourth quarter came in below expectations at $15 to $16 billion against $17.08 billion expected, which sent shares down about 20 percent in extended trading, while the client segment fell 5 percent to $8.1 billion.
On the product side, Intel confirmed that Nova Lake, its next client CPU generation, will be the first client chip built on the 18A-P process node. The company said the product is tracking ahead of schedule, and that expected design wins are roughly 2x those of Panther Lake, the current generation. Full technical and product details are scheduled for CES in January.
The 18A-P node brings measurable gains over the prior process: either +9 percent speed at iso-power, or -18 percent power at iso-speed, depending on which target a customer optimizes for. Intel also disclosed a sharp mix shift. 18A-based parts now account for about one third of Intel's client CPU revenue, compared with under 5 percent in Q1 2026, a ramp from a small base to a mainstream share in roughly two quarters.
Operationally, Intel said it is moving some Panther Lake production capacity toward data-center parts, specifically Granite Rapids. The sequencing suggests the company is prioritizing the data-center demand it just reported over client volume in the near term, allocating scarce leading-edge capacity where current revenue growth is strongest.
IMPACT: Intel's process transition is now showing up in reported revenue, not just roadmaps, and data center is the primary growth engine.
For teams building AI and automation infrastructure, the practical read is more supply and more competition in server CPUs. Faster Granite Rapids availability and a credible 18A-P client roadmap shorten refresh cycles and add another source of x86 capacity alongside AMD and ARM-based options, which tends to pressure pricing and improve lead times for GPU-adjacent server fleets.
Sources: cnbc.com — tomshardware.com
Continue the conversation: https://t.me/AIAutomationArchitecture · Bluesky
Top comments (0)