
Weebit silicon-oxide ReRAM headed to 28nm, AI
The company said it would target both CMOS bulk and fully-depleted silicon-on-insulator (FDSOI) manufacturing processes at the 28nm node and said migration could take place later in 2018.
The company also said it is on course to demonstrate 40nm 1Mbit memory array by mid-2018 and that the extended development program will ensure the technology is able to meet and exceed memory industry standards for performance and reliability.
Weebit Nano and Leti initiated their joint development program in September 2016 and the next phase of development will prepare the technology for a move to a commercial wafer fab.
Weebit said it has been receiving strong interest from industry players on the use of silicon-oxide ReRAM in the artificial intelligence field, such as deep learning and neuromorphic applications, and that it will continue to assess the opportunity.
Related links and articles:
News articles:
Weebit scales SiO2 ReRAM to 4kbit
Weebit moves SiOx ReRAM on to 40nm
University College London spins out ReRAM startup
Leti to develop SiOx memory with Weebit Nano
