After announcing its first quarter results for 2026, Intel disclosed that its foundry business is continuing to see positive changes. Among them, Intel 4, Intel 3 and 18A have entered key process nodes for mass production, and their yield rates have improved. This also means that the manufacturing foundation supporting most of Intel's current product portfolio is further improving.
Reports pointed out that Intel’s overall performance this quarter was strong, and the operating conditions of Intel’s product division and Intel’s foundry division have been optimized. In contrast, the progress of the foundry business has attracted particular attention because the unit has achieved yield improvements at all major mass production nodes simultaneously, which is regarded as an important milestone.
At the first quarter 2026 earnings call, Intel Chief Financial Officer David Zinsner said that the company is still continuing to improve the yield rate of more mature nodes such as Intel 4 and Intel 3. It is also continuing to optimize the current strongest performing 18A node to reduce manufacturing waste and increase the number of available chips, especially the yield performance on large-size chips.
According to David Zinsner, Intel's foundry business had an operating loss of US$2.4 billion in the first quarter, an improvement of US$72 million from the previous quarter. This was mainly due to the improvement in the yield rate of Intel 4, Intel 3 and 18A nodes, which promoted the improvement of gross profit margin. However, this improvement was largely offset by higher operating expenses, mainly related to Intel's intention to increase investment in the 14A node to support internal and external customer evaluation efforts.
He also said that Intel's foundry currently bears most of the costs in the early ramp-up stage of the 18A process. As 18A gradually moves towards larger-scale mass production and the yield rate continues to improve, Intel expects that the operating losses of the foundry business will continue to improve this year. In addition, Intel foundry's actual output this quarter was higher than the company's expectations, the yield rate maintained a steady improvement, and several key milestones for the 14A node were completed.
From an industry observation perspective, this shows that Intel’s plan to promote the recovery of the foundry business is gradually on the right track at least at the manufacturing execution level, and yield improvement is only part of it. The report also mentioned that Intel has won Tesla as the first large customer of the 14A node, but the relevant cooperation details are still limited.
In addition to yield improvement, another key task for Intel is to continue to update the process design kit, or PDK, for the 18A and 14A nodes. Reports show that the 14A node is currently in the 0.5 version PDK stage. Customers will finalize the mass production scale, design plan and other specific requirements after the 0.9 version PDK is released. Therefore, Intel still has time to further polish the node.
Reports indicate that Intel is currently maintaining collaboration with customers and continues to incorporate customer feedback into the node design optimization process. As 18A further ramps up volume and 14A development continues, whether Intel's foundry business can reduce losses through simultaneous improvements in yield, customer resources, and process maturity will remain the focus of the market.