It's hard to say when things turned dark for Intel or attribute any particular business decision that led to this situation, but yet here we are: Intel is bleeding money, the performance of its products is behind offerings from AMD and Nvidia, its foundry efforts yet have to take off as its current process technologies are behind products fabbed at TSMC, and it has to lay off tens of thousands of employees to cut costs and attempt to break even in 2026. Intel started to work on 20A and 18A technologies well before the company outlined its IDM 2.0 strategy in early 2021, so while the technology has a decent process design kit (PDK), is compatible with industry-standard EDA tools, and standardized IP blocks (e.g., SRAM macros, I/O cells, PLLs), it may still be more optimized for Intel's client and data center products rather than for a broad set of applications used by external clients. To add to an already rather intense situation on Intel's board, it should be noted that TSMC had little (if any at all) interest in taking over Intel's manufacturing facilities both because these are low-mix/high-volume fabs tailored for making CPUs and because the companies use completely different process technologies and converting Intel's fabs to TSMC's production methods wouldn't make financial sense for the company.
Author: ashilov@gmail.com (Anton Shilov) , Anton Shilov
Published at: 2025-08-11 20:57:12
Still want to read the full version? Full article
