The Chip Insider®–Intel Direct Connect 2025: Lip-Bu Tan’s Strategy for Intel

The biggest change since last year’s Direct Connect is that Intel has gone from knowing what a foundry is to comprehending how to be a foundry. I’m guessing you thought I would write that the biggest change was in management from Intel insiders to outsiders. No doubt, this was a big change. But it would have meant nothing if the company had not changed Gelsinger’s positioning, which reflected Intel’s historic IDM ethos. This year, there was no mention of IDM 2.0… “System Foundry”… was explained more clearly: with Advanced Packaging folding into the suite of services a foundry must offer… The biggest indication that Intel comprehends how to be a foundry was in how the need for Intel to earn the trust of its customers was now center stage...

Everybody wants to know what Lip-Bu Tan’s strategy is. On reverse-engineering, his public positions seem clear: Close Intel’s trust gap from its IDM days when the Wintel oligopoly gave it dominance over the world’s most profitable silicon real estate: PC CPUs. That’s changed. PC CPUs have become commodity parts…

The four parts in … earning trust … are … predictability, affordability, quality, and velocity… As for raw technology advantage, it is doubtful Intel needs to change anything. 18A has put them in the lead with TSMC choosing the more conservative route of splitting its Back-Side Power (BSP) introduction across two nodes. Some didn’t understand my comments about TSMC swapping roles of being the conservative leader with Intel...Back in the 1990s, TSMC took major manufacturing risks, like being the first to introduce minienvironments…switch from aluminum to copper interconnect...As time progressed and TSMC got larger, it became more conservative…

The reason why 18A offers a differentiable advantage is the gains in density, area, and cell utilization from BSP. If BSP is split across two nodes, all the IP validation from the first FSP design has to be redone to get these advantages (Front-Side Power)…The difference in risk-taking may be due to the different BSP approaches…These differences at the 20 Angstrom node could well decide who wins the next cycle of the foundry wars.

“If you don't like what's being said, change the conversation” – Don Draper, Mad Men