Nantero have been in the news recently following their successfully $10M Series D round of financing. Nantero was started in 2001 by three co-founders from Harvard University and Reed College. It was good to hear the company was still in business as they have, as far as I know, a unique approach to next generation memory resistive memory (called NRAM) based on a BEOL cell containing carbon nanotubes (CNT). There has not been much news from Nantero (based in Woburn, MA) for some time although the CTO, Thomas Rueckes, made an interesting presentation at the 2011 Flash Memory Summit (see recent Blogs for a discussion of presentations from the 2012 FMS).
The idea behind Nantero’s NRAM is that the nanotubes act as nano mechanical switches which close (SET) under electrostatic attraction induced by the application of a small electric field. RESET is phonon (heat) induced. The low resistance state has a very high activation energy (~5eV) attributed to physical adhesion (Van der Waals force). See above for a schematic cartoon from Nantero. More details at http://www.nantero.com/mission.html and the FMS 2011 presentation. A consequence of the high activation energy is that retention of the set-state is phenomenal, >1000 years at 85C and >10 years at 300°C being quoted. Presumably this is extrapolated data from Arrhenius type plots. Impressive endurance numbers are also presented. Data appears to be based on a high yielding 4Mb test chip fabricated with 140nm technology in a 1T-1R configuration.
The technology has migrated to an ‘in via like’ implementation similar to other ReRAM implementations although the memory layer is very different. In fact the technology is described by Nantero as “NRAM: RRAM with CNT Resistance Change Material”. The best image I found is not in one of the Nantero links above but at EETimes www.eetimes.com/electronics-news/4403877/10-tech-startups-to-watch-for-in-2013 where Nantero is identified as one of the top ten start-ups to watch in 2013. SET/RESET is fast so not unsurprisingly, Nantero see their technology as suited to both NAND and DRAM replacement as well as embedded NVM applications. Furthermore they have developed a transferable BEOL process including the CNT containing material.
Going back to the recent Series D funding, the round is described as being led by two new strategic corporate investors “currently engaged in strategic development and partnerships with Nantero”. While the two new investors are not identified, the Nantero website does haves a section entitled ‘Partnerships’ where 7 partners are listed. In addition, IMEC recently (10/31/12) announced a joint program with Nantero for CNT memory below 20nm. So what can we learn from all this? Given the negative reports on Nantero over the years it is certainly testament to the founders’ determination (and vision) that Nantero is still in existence and apparently thriving. And it is certainly remarkable that IMEC have not just teamed up with Nantero but have also publicly endorsed the promise of Nantero’s approach. However, I can see scaling issues related to the CNT deposition as multiple CNTs appear to be needed within each NRAM cell. Nonetheless, it will be interesting to see what comes out of the IMEC/Nantero project.
Christie Marrian, www-ReRAM-Forum Moderator