Looking at some of the smaller companies engaged with CBRAM and ReRAM, Blog sponsor Adesto Technologies is an obvious place to start. Teamed up with Altis semiconductor, Adesto have published interesting data on CBRAM integrated with CMOS in working memory arrays. We plan to return this development in a future Blog entry. Unity Semiconductor is one of the true veterans of the ReRAM field. However, they have been through some tough times recently with a major re-organization in 2010 followed by an agreement to with Micron the following year. Earlier this year it was announced that the memory company, Rambus, had acquired Unity. Nonetheless, Unity’s 64Mb test chip described at the ISSCC in 2010 remains one of the larger ReRAM memories fabricated to date and the accompanying paper includes a discussion of the challenges in sensing and programming of cross point arrays.
Crossbar Inc is a more recent ReRAM start-up that has come out of the University of Michigan. Information on the company is a little hard to find but they are building up a presence, particularly with a patent portfolio. Based on the co-founder’s publications in the open literature, the materials used are mostly Si based with an Ag electrode. Again it appears the memory array can be fabricated in the BEOL over CMOS. A similar approach is being pursued by 4DS, a start-up with provenance in Australia. A few details appeared in 2011 including 4DS membership of SEMATECH’s Materials and Emerging Technologies program. 4DS also announced a partnership with A*Star to fabricate a 16kb ReRAM. Another Australian spin-out is Qs who appear to have a different approach based on SiC however it is not clear to me at least if this is an ReRAM approach. Monolithic 3D are a 3D integration company who have some intriguing cartoons describing a 3D ReRAM at their website based on the ‘smart cut’ process which can be used to produce SOI wafers. I can see challenges but the end result is claimed to be a transistor structure associated with each memory element.
I suspect that these smaller companies will be the source of the first ReRAM/CBRAM products. Almost all the larger companies have the problem that initially at least an ReRAM/CBRAM chip will compete with some of their own products. A start-up does not have that problem. On the other hand, one of their main challenges is to identify a particular application where the particular characteristics of CBRAM/ReRAM technology provide an effective solution.
Christie Marrian, WWW.ReRAM-Forum.com Moderator