PORTLAND, Ore. -- MRAM speed can be boosted and power consumption cut by using a new electrical switching technique that requires much less energy to write bits, according to Chinese researchers.
The basic idea is to switch magnetic domains "halfway" rather than to completely reverse their magnetic orientation, which the researchers at Tsinghua University (Beijing) claim still enables MRAMs to store binary bits, but at much faster switching speeds and using a fraction of the energy normally required.
Traditional MRAMs use magnetic fields to switch the state of bit cells, resulting in poor densities compared to flash. This week, Japanese researchers described how MRAMs could utilize electrical switching to perform perpendicular recording that boosts MRAM densities past flash. The Chinese team claims that electrically switched MRAMs can also be made faster and more energy efficient than today's magnetically switched devices.
Unlike magnetically switched MRAM bit cells, which require a complicated multilayered stack, the Chinese researchers constructed their electrically switched MRAM bit cell using just two layers of different ferroelectric films. An electrical signal was shown to be capable of affecting the magnetic polarization of the bit cell by scattering the walls of the striped domain in the bilayered structure. This transformed the structure into a single domain that changed the film's resistivity just enough to detect.
The Chinese researchers demonstrated that applying an electrical voltage to their MRAM bit cell allows the presence or absence of domain walls to be used to store information. They are currently optimizing their material stack for commercialization by enhancing the resistivity change induced by the presence or absence of domain walls.
Thanks for the link to the publication littletiger.
In the reference list of this one, I will try to locate (and post) a publication referring to the Japanese team who demonstrated how electrical switching could be utilized to perform perpendicular recording that increase MRAM densities.
By the way, can some one please point out a direct link to the latter publication from Japanese team?
True, the article skips over quite significant technical details that are of basic interest to EEs. However, the combined research outcome that has been presented here appears to be of substantial value.
Therefore, could someone please provide links to one or two of the key research papers that outlines these two discoveries?
Looking at this article from a different point of view, it is quite interesting to note how the global research happen in these modern times. On one hand, a Japanese team demonstrates how MRAMs could be switched electrically to increase density beyond what has been possible so far. On the other hand, a Chinese team declares how electrically switched MRAMs could be made faster and more energy efficient. Once proven, these two technologies combined, presents quite an opportunity for a breakthrough memory product. Therefore, the concept could prove to be of interest not only to electrical engineers, but also to investors and venture-capitalists alike.
Another point of interest, in these first few months of the post-recession area, is on how novel technologies that are completely based in the far-east come into being. Earlier, this was not the case as US and the West had some part to play in the majority of these. Is it not that the far-east appears to come out stronger after the recession?
David Patterson, known for his pioneering research that led to RAID, clusters and more, is part of a team at UC Berkeley that recently made its RISC-V processor architecture an open source hardware offering. We talk with Patterson and one of his colleagues behind the effort about the opportunities they see, what new kinds of designs they hope to enable and what it means for today’s commercial processor giants such as Intel, ARM and Imagination Technologies.