02/14/2025, 04:55 AM UTC
台积电2nm密度更高，但Intel 18A性能更强？TSMC's 2nm Density Higher, But Intel 18A Performance Stronger?
➀ 半导体研究机构TechInsights和SemiWiki在“国际电子设备会议”上公布了英特尔和台积电即将推出的18A（1.8nm级）和N2（2nm级）工艺技术的关键细节。
➁ 英特尔18A提供更高的性能，而台积电N2可能会提供更高的晶体管密度。
➂ 与上一代3nm（N3E）节点相比，台积电N2制程在相同电压下可以将功耗降低24%至35%，而英特尔18A在2nm级工艺中具有最高性能。
➃ 台积电N2的高密度逻辑单元晶体管密度为313 MTx/mm2，英特尔为238 MTx/mm2，三星为231 Mtx/mm2。
➄ 台积电N2的256Mb SRAM阵列的平均良率为>80峰值良率为>90%。
➅ 英特尔18A计划于2025年年中投产，而台积电N2计划于2025年底投产。
➀ TechInsights and SemiWiki revealed key details about Intel's 18A (1.8nm) and TSMC's N2 (2nm) process technologies at the International Electron Devices Meeting (IEDM).
➁ Intel 18A offers higher performance, while TSMC N2 may provide higher transistor density.
➂ TSMC N2 can reduce power consumption by 24% to 35% compared to the previous generation, while Intel 18A has the highest performance among 2nm process technologies.
➃ TSMC's N2 has a higher density of high-density logic unit transistors, with a density of 313 MTx/mm2, followed by Intel with 238 MTx/mm2 and Samsung with 231 Mtx/mm2.
➄ TSMC's N2 is expected to have an average yield rate of >80% for 256Mb SRAM arrays, with a peak yield rate of >90%.
➅ Intel 18A is scheduled to enter mass production in mid-2025, while TSMC's N2 is planned for late 2025.
---
本文由大语言模型（LLM）生成，旨在为读者提供半导体新闻内容的知识扩展（Beta）。
