Intel Ohio factory construction delayed, AEP’s $95 million substation idled. Fujitsu 2nm CPU is manufactured by TSMC, but it will consider using Rapidus to ensure a stable supply chain. Counterpoint Research: 3nm and 2nm will account for one-third of smartphone…
Samsung recently postponed the construction of a 1.4nm test line that was originally scheduled to start in Q2 2025 and will focus all its attention on 2nm. Sedaily Tesla is stepping up its investment in AI and will invest $8 billion…
The first chip of the UALink camp is expected to be taped out at the end of 2025 and will challenge Nvidia. Counterpoint Research: Global semiconductor wafer foundry 2.0 market revenue was US$72.29 billion in Q1 2025, a year-on-year increase…
TSMC is actively planning 2nm production capacity, and it is expected that the monthly production capacity will reach 200,000 pieces by the end of 2028. Samsung foundry lost a customer, and Google Pixel 10 phone 3nm chip orders will be…
Intel 18A process faces strong challenges from TSMC 2nm process, with a yield rate of only 20-30%. Foxconn will use humanoid robots to produce Nvidia AI servers, which may be put into operation by Q1 2026. It is rumored that…
It is reported that SoftBank plans to invest $1 trillion to build an AI and robot manufacturing park in US and hope TSMC will invest and join this project. Foxconn and Nvidia team up to break the deadlock: humanoid robots…
Samsung Electronics will provide AI education to 100,000 students. Although NVIDIA has opened NVLink licenses, it still sets many software and hardware restrictions. DigiTimes It is reported that Tesla’s AI5 chip based on TSMC’s N3P process has entered the mass…
Nvidia will launch a new RTX 5090DD graphics card for the Chinese market to replace the 5090D. Intel director: Future chip manufacturing will rely more on etching than lithography. SK Hynix has already supplied HBM4 to Nvidia in small batch…
Intel 18A will go into mass production at the end of 2025. TSMC global expansion is also related to Taiwan’s “power shortage”. Marvell raised its 2028 total addressable market (TAM) forecast for its custom AI chips to $55 billion. Morgan…
Intel Lip-Bu Tan appointed semiconductor industry executives Srinivasan Iyengar, Jean-Didier Allegrucci and Shailendra Desai to lead customer engineering and AI chip development. MiniMax released its first open source inference model, MiniMax-M1, which cost only $530,000 to train. Texas Instruments announced…
