According to Infineon, its next-generation Aurix microcontroller will use embedded non-volatile memory, specifically resistive random access memory (RRAM), instead of embedded flash (eFlash), and it will be manufactured on TSMC's 28nm node.
While samples of the Autrix TC4x family of microcontrollers based on TSMC's 28nm eFlash technology have already been shipped to lead customers, the first samples based on TSMC's 28nm RRAM technology will be available to customers by end of 2023. Infineon said that the Autrix TC4x famility of microcontrollers are designated for ADAS, and offer new E/E architectures and affordable AI applications.
Embedded flash microcontrollers have been used as ECUs in automobiles since the introduction of the first engine management systems, and the majority of MCU families in the market are based on eFlash technology, but the technology have been struggling to migrate below 28nm, and it is also considered less efficient than RRAM. Infineon believes that the cooperation with TSMC has successfully created the basis of RRAM in the automotive domain, and put its Autrix family of microcontrollers on a broader supply basis.
Kevin Zhang, TSMC's senior vice president of business development, stated that TSMC and Infineon have already partnered on RRAM technology for almost a decade in a range of different applications, and moving the TC4x to RRAM will open new opportunities in terms of shrinking microcontrollers into smaller nodes.
Currently, TSMC's non-volatile memory solutions include Flash, spin-transfer torque MRAM (STT-MRAM), and RRAM. The foundry is also exploring phase change RAM (PCRAM), and spin-orbit torque MRAM (SOT-MRAM) technologies.
In 2018, TSMC began the volume production of 40nm eFlash technology for automotive, but its 40nm ultra-low-power embedded RRAM technology, fully compatible with CMOS process, already entered risk production at the end of 2017. In 2021, the foundry's 40nm RRAM technology successfully entered volume production, with 28nm and 22nm nodes also available as a low-cost solution for IoT market.