Non-Volatile Memory (NVM) includes the industry common One-Time-Programmable memory (OTP), Multiple-Time-Programmable memory (MTP), Flash memory (Flash), and next generation NVM of Magnetic RAM (MRAM) and Resistive RAM (RRAM).
TSMC provides foundry's most advanced and comprehensive portfolio of Embedded NVM technologies, featuring fastest computing capacity, smallest flash dimensions, and lowest power consumption.
TSMC's Embedded Flash technology ranges from 0.5-micron (µm) to 40nm and provides a number of flash IP options to meet a variety of product design requirements. TSMC's 2018 total shipment of Embedded Flash wafers to customers ranked No.1 in the global semiconductor industry, supporting a wide range of applications including consumer electronics, portable electronics, home appliances, smartcards, computers, wired and wireless communications, automotive electronics, medical equipment, industrial automation, the Internet of Things, and wearables.
Meanwhile, TSMC started volume production of 40nm Embedded Flash technology for automotive in 2018, and is now developing the 28nm Embedded Flash. This technical qualification is expected in 2019 for automobile electronics and micro controller units (MCU). At the mean time, TSMC is also developing Embedded MRAM, and Embedded RRAM in parallel to fulfill customers' need of continuous performance improvement and power-consumption reduction. 40nm ULP embedded resistive random access memory (RRAM) technology, which began risk production at the end of 2017, completed consumer grade qualification test for 10,000 cycles of endurance in 2018. This technology is fully CMOS (Complementary Metal Oxide Semiconductor) logic compatible for PDK and IP re-use for applications including wireless MCU, IoT and wearable devices. 22nm ULL magnetic random access memory (MRAM) technology progressed well, demonstrated reflow capability and passed JEDEC 168 hours high-temperature operating life (HTOL) reliability validation at the end of 2018. Through IP customization, MRAMs can serve various applications, such as artificial intelligence and eFlash replacement for MCU.