Intel has spent the previous couple of years making an attempt to reinvent itself and show it’s nonetheless related in an AI-centric world dominated by Nvidia’s chips.
On Thursday, as Intel crushed Wall Road monetary targets, the corporate had a brand new message: There’s nothing fallacious with being a 58-year-old maker of PC and server microprocessors.
“We are embracing our roots as data driven, paranoid, and engineering driven,” CEO Lip Bu Tan mentioned initially of the corporate’s Q1 earnings convention name, referencing the well-known “only the paranoid survive” philosophy of Andy Grove, the late cofounder of Intel.
Shares of Intel surged greater than 22% in after hours buying and selling Thursday after the corporate reported first-quarter outcomes. As a substitute of the two% lower in income that analysts have been anticipating for the primary three months of the yr, Intel grew income 7% year-over-year to $13.6 billion. Income within the present quarter will vary between $13.8 billion and $14.8 billion, Intel mentioned, effectively above the $13.06 billion analysts have been anticipating.
Demand for Intel’s central processing items (CPU) chips, that are based mostly on its longstanding x86 structure, is booming, the corporate mentioned. Actually, income would have been even larger had it been capable of produce extra of the chips.
“A year ago the conversation around Intel was about whether we could survive,” Tan mentioned. “Today it’s about how quickly we can add manufacturing capacity and scale our supply to meet enormous demand for our products.”
It was hardly an exaggeration in the case of the grim outlook for the corporate, which he joined as CEO in March 2025, a number of months after Pat Gelsinger was ousted from the highest job. On the time, many observers, together with former board members, questioned whether or not the corporate needs to be damaged aside, with its manufacturing amenities bought or spun right into a separate enterprise. A couple of months after Tan began, the U.S. authorities purchased a ten% stake in Intel, serving to to shore up the corporate in a deal the Trump administration mentioned was vital for nationwide safety and American trade.
CPUs are again, however is Intel?
The resurgence in demand for Intel’s CPUs is a considerably stunning flip of occasions after a number of years by which the GPUs, or graphics processing items, made by Nvidia seemed to be the long run due to their prowess with AI fashions.
“In recent months we have seen clear signs that the CPU is reasserting itself as the indispensable foundation of the AI era,” Tan mentioned on the decision. The explanation, he defined, is that CPUs are higher suited to working AI providers, versus creating—or coaching—AI fashions, the place GPUs have the sting. Within the early days of the generative AI increase, as firms like OpenAI, Anthropic, and Google have been coaching big new AI fashions, GPUs have been the clear winner. However because the market evolves, Intel mentioned the pendulum is swinging again to CPUs.
Intel finance chief Dave Zinsner mentioned that the ratio of GPUs to CPUs in AI information facilities is altering. Whereas there are usually seven or eight GPUs for each one CPU for the job of coaching AI fashions, the ratio is simply three or 4 GPUs for each one CPU in the case of inference, or working AI fashions. And as agentic AI positive factors floor, Zinsner mentioned the ratio may hit parity and even flip in Intel’s favor.
However there are nonetheless loads of challenges. Nvidia lately launched its first standalone CPU, including to current competitors Intel faces from longtime rival AMD, in addition to from server chips based mostly on the ARM structure (together with an upcoming chip that ARM is making itself, as a substitute of strictly licensing the chip design to different firms).
And the larger query is whether or not Intel’s resurgence is really an indication that the corporate is on the mend, or just a mirrored image of the booming AI infrastructure buildout, as information heart firms snap up as many chips as they’ll. Massive questions additionally stay about Intel’s so-called foundry enterprise, which manufactures chips for different firms and competes with international big TSMC—notably whether or not Intel will proceed to speculate the huge sums required to develop the following technology of chipmaking know-how.
Tan has beforehand mentioned Intel wouldn’t decide to constructing factories utilizing essentially the most superior 14A fabrication course of (able to producing chips with 1.4 nanometer circuits) except it has dedicated prospects. And he gave no replace on that entrance on Thursday, regardless of hypothesis that Elon Musk and Telsa’s lately introduced partnership with Intel, through Terafab, could be the much-anticipated 14A buyer.
Requested about Terafab deal, Tan described it as a broad relationship by which the 2 firms will be taught lots collectively, however offered few specifics. “Elon and I believe the global supply chain is not keeping pace with the rapid acceleration in the demand,” he mentioned.
As for 14a prospects, Tan was equally tight lipped: “We’re making great progress in terms of yield and cycle time. And clearly we’re engaging with multiple customers; heavy engaging. My style is underpromise, over delivering. So we have no plans to announce the customer unless a customer wants to announce it.”
