Lip-Bu Tan says bypassing management silos to read engineering reports directly fixed broken yields and simplified a chaotic product roadmap.
The Update: Intel CEO Lip-Bu Tan revealed that upon taking the helm, he mandated all engineering teams report directly to him to eliminate corporate silos. This structural bypass allowed him to personally diagnose failures in the 18A manufacturing node, simplify a bloated product roadmap, and achieve a seven to eight percent monthly yield improvement—a metric he claims is now hitting industry-best standards. (Source: PC Gamer / Comcast MSNBC interview, May 20, 2026)
Breaking the Silos
In a recent interview with Jim Cramer, Tan detailed a blunt management reset. His predecessors allowed engineering data to bottleneck before reaching the executive suite. Entity → Mechanism → Outcome: [Intel Engineering] → [Direct CEO reporting line] → [Immediate diagnosis of manufacturing bottlenecks].
When Cramer asked if engineers reported to the previous CEO, Tan simply said, "No." He cited a culture with "too many silos, too many people needed reporting," where critical technical issues were diluted by the time they reached the top.
Why was the 18A manufacturing yield failing?
Tan inherited a struggling Intel Foundry. The 18A process node—crucial for competing with TSMC and Samsung—suffered from poor manufacturing yields. (Manufacturing yield measures the percentage of functional chips produced per silicon wafer; low yields mean wasted materials and uncompetitive costs).
Rather than relying on internal filtered summaries, Tan brought in ecosystem partners to audit the raw data. Entity → Mechanism → Outcome: [18A Process Node] → [Ecosystem partner data auditing] → [7-8% monthly yield recovery]. This external benchmarking forced internal accountability.
The Roadmap Purge
Beyond manufacturing metrics, Tan explicitly stated the previous product strategy was a mistake. "We made a lot of mistakes, and now we correct that mistake, and we simplify the roadmap."
By holding the engineering reports himself, Tan aimed to "focus on the engineering, how to redesign, [how to] simplify the products, and then get the real killer product out." Entity → Mechanism → Outcome: [Product Roadmap] → [Reduction of overlapping architectures] → [Targeted "killer" product development].
What this means for the market
For hardware watchers and enterprise buyers, this shift is significant. Tan isn't just a financial steward; he is functioning as a technical gatekeeper.
- For Competitors (TSMC, AMD): A streamlined Intel Foundry with transparent yield metrics poses a credible threat in contract manufacturing.
- For Ecosystem Partners: Direct data-sharing with the CEO suggests Intel is moving faster to fix flaws, potentially shortening the time from prototype to mass production.
Entity → Mechanism → Outcome: [Intel Foundry] → [Yield normalization] → [Viable alternative manufacturing hub].
What is still unknown
While Tan claims 7-8% monthly yield improvements, actual independent yield metrics for the 18A node remain confidential. The exact products on the "simplified" roadmap haven't been fully detailed to the public, leaving enterprise clients to wait for official architecture briefings.
What to watch next
- Next Intel Architecture Briefing: Look for confirmation of canceled or merged product lines.
- Foundry Client Announcements: External companies signing on to use Intel 18A will validate the yield claims.
- Engineering Turnover: Monitor whether direct reporting to the CEO is sustainable long-term or causes burnout in senior engineering tiers.
