Until now, emerging non-volatile memory (eNVM) like spin-transfer torque magnetic RAM (STT-MRAM), resistive RAM (ReRAM), and phase-change memory (PCM) have mostly been restricted to niche markets because of poor scalability, high cost, and a lack of support from major memory makers. However, the 2017 introduction of Intel’s Optane products, based on 3D XPoint (a technology jointly developed by Micron and Intel) can be viewed as the defining moment for the adoption of emerging NVM in mainstream applications.
In the stand-alone business, PCM is the leading emerging technology thanks to the increasing adoption of 3D XPoint as an enterprise and client storage-class memory (SCM). In this context, Intel has a unique position amongst IDMs: it is a stand-alone memory supplier and a CPU leader, and it can combine its new persistent-memory modules (Optane DIMMs) with its new generation of Xeon processors (i.e. Cascade Lake), which will act as a Trojan horse for introducing 3D XPoint into the data-center technology market – thereby boosting its sales. It is worth noting that Samsung, Toshiba, and Western Digital took a different strategic path by developing 3D NAND-based SCM solutions like Z-NAND (Samsung), XL-Flash (Toshiba), and Low-Latency (LL) Flash (Western Digital). However, these technologies will be used in enterprise SSDs and thus will not compete with Intel’s new Optane DIMMs (soon to be globally available), which we expect will represent more than 50% of overall 3D XPoint sales.
Compared to stand-alone, the embedded eNVM market is relatively small, representing just ~2% of the total eNVM market. In 2018, embedded ReRAM contributed to the majority of revenue, thanks to the sales of low-density ReRAM-based MCUs (i.e. Panasonic). However, all the top foundries are involved in STT-MRAM with 28/22nm technology processes, whereas RRAM has encountered some delays. STT-MRAM is gaining strong momentum for applications in MCUs, SoCs, and other ASIC/ASSP products. Recently, Samsung commenced commercial production of 28-nm embedded STT-MRAM, and it also announced plans to tape out a 1 Gb test chip before the end of 2019. At the same time, Intel has confirmed that its embedded STT-MRAM has reached high yields suitable for mass production. Moreover, embedded PCM is not yet out of the race either – STMicroelectronics is currently developing PCM-based MCUs for the automotive industry. Indeed, 2019 is expected to be rich with new events and developments for eNVM.
In the “Emerging Non Volatile Memory 2018” report, Yole has deeply looked into the evolution of emerging memory technologies and markets, providing a comprehensive analysis of the players’ roadmaps and dynamics in the growing eNVM business.
Simone Bertolazzi, PhD is a Technology & Market analyst at Yole Développement (Yole) working with the Semiconductor & Software division. He is member of the Yole’s memory team and he contributes on a day-to-day basis to the analysis of nonvolatile memory technologies, their related materials and fabrication processes. He (co-) authored more than ten papers in high-impact scientific journals. Simone obtained a PhD in physics in 2015 from École Polytechnique Fédérale de Lausanne (Switzerldand), where he investigated novel flash memory cells based on heterostructures of 2D materials and high-k dielectrics
After more than 15 years in development, PCM has finally taken off in stand-alone applications. STT-MRAM will lead the embedded memory race.