➀ TSMC is expected to have the best 2nm process by the end of the year, with 60% yields on a GAA process in mass production in Q4 2025. ➁ Samsung is aiming for a Q4 launch for production on 2nm but won't optimize BSPD until 2027. ➃ Intel is also targeting a 2nm process with BSPD and GAA, despite 10% yields reported. ➄ Rapidus, a dark horse, is hoping to have its first 2nm prototype chips out this April with production by the end of the year and mass production in 2027.

➀ Rapidus has received an ASML EUV machine for installation at its IIM-1 foundry in Chitose, Hokkaido; ➁ This marks the first time EUV technology will be used for mass production in Japan; ➂ The TWINSCAN NXE:3800E machine can process 220 wafers an hour; ➃ Rapidus plans to start pilot silicon production at IIM-1 in April 2025 with a single-wafer process.

➀ Intel's initial promises under Pat Gelsinger were ambitious but execution has been lacking; ➁ The company faces challenges with a bloated workforce, risk-averse culture, and a lagging AI strategy; ➂ Intel's investments in process technology have not yielded significant results, and the company must make hard choices about its future.

➀ TSMC will start producing 2nm GAA-based wafers in Q4 2025 at Fab 20 in Hsinchu; ➁ The production will be followed by Fab 22 in Kaohsiung in Q1 2026; ➂ N2P production will start in late 2026 without backside power delivery; ➃ N2 and N2P will use TSMC's NanoFlex technology; ➄ N2 is 15% denser than N3E and improves performance by 10% to 15%; ➅ N2E will be followed by N2P and N2X in 2026; ➆ TSMC's planned capacity for 2nm is more than for 3nm; ➇ 2nm wafers are expected to cost $30k+.

➀ Intel has made a strong move after poor Q3 results, repositioning its foundry unit as an independent operation. This is aimed at attracting US rivals like AMD, Nvidia, and Qualcomm. The US Department of Commerce is encouraging this shift, with half of the world's top ten fabless companies based in the US. ➁ Intel's foundry unit is seen as a viable option for US hyperscalers looking for proprietary SOCs. ➂ With Intel's promise of the world's best foundry process next year, there's potential for significant growth.

➀ Intel's ability to catch up with TSMC in process technology is questioned; ➁ Malcolm Penn, CEO of Future Horizons, suggests it could take 10 years to catch up; ➂ Intel's process capability is about a year behind TSMC; ➃ TSMC has been running its 3nm process for a year, while Intel's best process in volume production is Intel 4 at Leixlip.