When you buy through our articles links, Future and its syndication partners can earn commissions.
Intel was late in all means. Borked processors, failed FABS, cash bleeding, start with US president. The troubles never end. Until next year, that is, Nova Lake launch. That’s right, peeps when Intel’s next generation CPU architecture will appear in 2026. At the end of the year, Intel will rise to a leading position across the desktop.
Quoting John Pitzer, Vice President of Intel Corporations related to corporate planning and investor relationships, exactly: “When Nova Lake will appear at the end of next year by 2027, I think we will have a leadership on the desktop.”
This is one of the bold claims, even because of the chip, which has rumored to pack up to 52 kernels. Of course, at least partly will depend on the impending Intel 18A process quality – a node, formerly known as the one who is also Gelsinger betting the entire company, but later more steps to profitability, and 14a is a knot that Intel expects to create customers now customers.
Not surprisingly, Pitzer is also 18a. He says Intel plans to spend a lot of tools to get a lot of chips 18A and return the Intel profitability. Sounds, or you can say that both of the latest Intel CPU families, namely Arrow Lake and Moon Lake, are mainly produced by TSMC.
Intel will begin the process of transferring this production home with Panther Lake. However, the Panther Lake is a mobile processor and is also designed to use energy more efficient laptops than larger desktop replacement platforms. When it comes to this, it is believed that only in certain versions of Panther Lake will feature Intel’s processor stamps. Most of the production can be completed by the Taiwan chip foundry TSMC.
In other words, Panther Lake will only have a relatively limited amount of Intel 18A silicon. This changes with Nova Lake. “Nova Lake itself is both a laptop and a part of the desktop has a fairly significant effect on the amount of waffle starts we need 18A,” says Pitzer.
And it should mean improved profitability, as it means that Intel will not have to pay those expensive TSMC waffles. “When you look at it through the Intel Foundry lens, the transition from the Intel 7a Intel will increase three times faster than their price through the waffle. So just ride more volume via Fab on 18a is a pretty profitable Intel Foundry dynamics, “and Pitzer.
Morever, Pitzer says Intel can go back to money earning, simply by doing its products, he doesn’t need to win big customers. “We do not need to see a lot of external foundry income to give up in 2027,” he says.