Intel has significantly reduced its management hierarchy, cutting the number of vice presidents from approximately 450 to around 200. The company has also halved its management layers, decreasing from 12 to six. This restructuring is part of CEO Lip-Bu Tan’s efforts to enhance decision-making speed and ensure timely product shipments. CFO David Zinsner revealed these changes at the Deutsche Bank Technology Conference, characterizing the previous structure as a cultural issue that hindered Intel’s operational effectiveness for over a decade.
Intel had 108,900 employees at the end of 2024 and 88,400 by late September 2025, including subsidiaries like Mobileye. Around 20,500 of those exits happened under Tan, who took charge in March 2025 and told staff there would be “no more blank checks. Intel is now preparing for 14A risk production in 2027 and high-volume manufacturing in 2028. Intel entered 2026 expecting gross margins in the high 30s and is now in the low 40s. The target is the mid-40s, then the high 40s, and eventually a figure starting with 5. If it reaches risk production in 2027 and volume manufacturing in 2028 as planned, it will be the clearest evidence yet that the cultural reset worked. If it slips into the multiple development cycles that marked the last decade, the 250 fewer vice presidents will have changed very little.
Zinsner said too many people across the organisation held veto power, so decisions kept stalling. Looking back at the last decade, he said much of Intel’s trouble “can be boiled down to culture,” adding that 50-person startups were getting products out faster than Intel. Information changed as it moved up the management chain, and by the time presentations reached the CEO, they often described a different situation from the one teams lower down knew about, according to Zinsner. The pitch was about flattening management, but in Oregon only 8% of the laid-off employees had “manager” in their job titles. To get there, Intel uses an internal “Rule of 45” that combines revenue growth and operating margin.
The damage showed up in products. Chips often needed four or five development cycles before they were ready, when they should have reached production after the first attempt. Transparency was the other big issue. He did not frame this as deliberate deception. Some managers simply did not want to admit a project was failing. The result was that Intel’s top leadership sometimes made decisions based on an inaccurate picture of the business. Tan has since brought in executives he trusts who are used to working in leaner organisations. The VP cuts sit inside a much larger downsizing. Most were engineers and technicians. Zinsner pointed to early signs of change. Several products have reached A-stepping, Intel’s term for the first version of a chip, without the repeated development cycles that were once routine. Intel 18A yields are running ahead of internal milestones, and 14A defect density is tracking better than the company’s target curve. Internal product teams have started designing for the process, and interest from external foundry customers has picked up. Finances are moving too. Slower businesses must deliver stronger margins, while faster-growing ones can earn more investment. Halving the VP ranks gives Intel a leaner org chart, but the numbers Zinsner shared will only mean something if they hold. 14A is the first major process being developed entirely under Tan’s structure, with fewer layers between engineers and the CEO and fewer people who can block a decision. You use AI every day. Now get your AI Quotient. Take the AIQ test.