May 28

Patent Landscape by Blaise Mouttet

Blaise Mouttet is an interesting guy who runs a nanotech blog focused on IP  including ReRAM. He has just updated his study on US patents related to Resistance Switching Materials and Devices. There is a huge amount of analysis and information in this package. I can’t do it justice in this Blog but here are a few thoughts on the contents. Blaise’s study “include issued US Patents spanning from the 1960s until February 25, 2014…Primary focus of patents collected is PCRAM (phase change), ReRAM (metal oxide), CBRAM (ionic/electrolytic), and other non-volatile memory materials based on a resistance switching effect.” Numerically phase change dominates with ReRAM and CBRAM 2nd and 3rd.

Blaise has compiled a ‘league table ‘ of Patent assignees based on patents issued since 2003 and provides an analysis of their main focus. Samsung is the largest US patent assignee in the study with.” 587 patents ….(Samsung) possesses patent for almost every resistance switching material type (phase change, metal oxide, CBRAM, molecular/polymer) with a primary interest in PCRAM (275 US patents) and a secondary interest in metal oxide ReRAM (34 patents).” Micron is next with “490 US patents for resistance switching materials. Patents are almost equally divided between PCRAM with 112 patents and CBRAM with 132 patents along with a few recent carbon-based ReRAM patents” Toshiba is 3rd with “286 US patents for resistance switching materials … with 100 issued in 2013 alone. Most of Toshiba’s patents are not limited to a specific resistance switch type …. Primary application focus is on alternative solution to non-volatile memory with an emphasis on 3D stacking solutions”.

Further down the list we find IBM with “188 US patents for resistance switching materials, mostly focused on PCRAM (phase change) but some metal oxide ReRAM (10) and CBRAM (4). Application focus includes neuromorphic processing (6 patents) using phase change devices.” And HP with “135 US patents for resistance switching materials. Originally focused on molecular resistance switching but now focusing on metal oxide systems” encompassing a broad range of application areas. HP is just ahead of Panasonic with 121 US patents for resistance switching materials. Mostly focused on metal oxide ReRAM particularly tantalum oxide based. Primary application focus is on alternative solution to non-volatile memory.”

Blaise also lists the expiration date for patents held by the various assignees. Samsung appears in good shape in this regard but Micron, Ovonyx (PCRAM), Round Rock Research (PCRAM) and Axon (CBRAM) all have a number of patents expiring over the next few years. I’m no lawyer but it will interesting to see the strategy that Axon, in particular, employ with respect to their expiring IP.
Blaise1I extracted the charts from Blaise’s database that shows the number of patents granted per year in various material systems. I’ve left the original classifications but metal oxide is predominantly ReRAM and ionic/electrolytic, CBRAM. Phase change (PCRAM) is tailing off as ReRAM (and Si based materials) have shown a recent dramatic growth while CBRAM has been steadier over the past decade or so. I couldn’t resist adding the molecular/polymer patents which no longer appear to be in vogue following the mid 2000’s peak. The second chart is the same data only without the phase change data.Blaise2

Blaise3The third chart shows only patent assignees I identified as small companies or academic institutions again without PCRAM.

Thanks, Blaise for sharing this remarkable data set
Christie Marrian, ReRAM Forum Moderator


  1. Fred Chen

    I found the patent 4272754 filed in 1979, its first claim reads:

    A nonlinear heterojunction comprising in layered relationship:
    a thin film layer of a first metal
    a thin film layer of a first metal oxide
    a thin film layer of a second metal oxide
    a thin film layer of a second metal.

    I think that might describe a good number of the listed patents.

  2. admin

    Hi Fred
    I’m no patent law expert but I suspect that US4272754 may not be relevant in terms of ReRAM as no memory function is claimed. Also it may well have expired!.
    Blaise has an interesting Blog on his TinyTechIP site entitled The memory resistor patent thicket (4th article from top).that looks into the issue of overlapping patents. I expect they will be no resolution until the lawyers have been unleashed……

  3. Fred Chen

    It should be expired (more than 20 years), but it’s still necessary for the ReRAM or other memory structure to distinguish beyond this claim, in a demonstrable, non-obvious way.

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>