Intel Is Allegedly Choosing TSMC for 2nm Nova Lake CPUs, Indicating a Possible Foundry Realignment and Dual-Sourcing Shift
According to a Taiwan Economic Daily article, Intel has reportedly placed orders with TSMC for its next-generation Nova Lake CPUs, choosing to use the Taiwanese foundry's state-of-the-art 2nm (N2) production node in a calculated move that represents a larger shift in manufacturing strategy. For Team Blue, which has historically depended on its in-house Intel Foundry Services (IFS) subsidiary but now seems more dedicated to a dual-sourcing strategy, this is a big move.
Intel’s decision to outsource the computational tile of Nova Lake to TSMC shows a prioritization of time-to-market and competitive efficiency over complete internal vertical integration. The action, which was allegedly spearheaded under the leadership of Products CEO Michelle Johnston Holthaus, demonstrates a readiness to depart from established procedures in order to provide customers with best-in-class performance. Holthaus had previously stated that if using external foundries like TSMC meant better satisfying customer expectations, Intel would not think twice about doing so.
Top tech companies have swiftly turned the TSMC 2nm process into a battlefield. Apple is anticipated to employ 2nm for the A20 chip in its iPhone 18 series, while AMD is already on board as the first customer, including it into its next 6th Gen EPYC "Venice" processors. The addition of Intel to the roster serves to further solidify the node's standing in the upcoming silicon innovation wave. However, given the current uncertainty in the global supply chain, adopting TSMC's N2 comes at a high cost, both financially and logistically.
While Intel’s reliance on TSMC may seem like a step back from its in-house ambitions, it's worth noting that Intel Foundry is not being sidelined entirely. The company still intends to utilize its own 18A node for products like the upcoming Panther Lake SoCs and Clearwater Forest Xeon processors. The effectiveness and adoption of these internally manufactured chips could dictate the long-term role of IFS, particularly if Intel’s flagship consumer products increasingly lean on TSMC's superior yield and performance margins.
With reports suggesting configurations of up to 16 Performance Cores (P-Cores) and 32 Efficiency Cores (E-Cores), Nova Lake itself is anticipated to represent a significant advancement. In line with Intel's larger initiatives to redefine scalability and performance-per-watt in the desktop and mobile computing categories, this would essentially double the core counts of current designs if it were to become a reality.
This renewed commitment to dual-sourcing is in line with Intel's overarching goal of taking the lead in process technology once again by 2025. However, the company's primary goal in the short term is straightforward: make sure that the most competitive CPUs are delivered to the market, even if that involves contracting with a rival foundry to produce its most ambitious chips.
Do you see Intel’s dual-sourcing strategy as a sign of flexibility and innovation, or does it suggest a loss of faith in Intel Foundry’s roadmap? Join the debate below.
