It’s that time of year again… Time to dust off that ol’ crystal ball and gaze into the murky depths and come up with the ReRAM-Forum predictions for the rest of 2014!
1) Mass Storage: No standalone ReRAM memory chips in 2014. I said the same last year and I see no reason that 2014 will be any different. The reasons were put most succinctly by David Eggleston in his FMS presentation in August. There is still life in planar NAND, 3D NAND is next, has already been released by Samsung and is on all the memory chip manufacturers roadmaps.
2) ReRAM memory: The first ReRAM memory chip to reach volume production will be the result of the Sony/Micron ReRAM collaboration. Sony has described a 16Gb chip at FMS and there is a talk about what appears to be the same chip (27nm technology) at ISSCC in February. Sony has mentioned several times that this will be available in 2015. 16Gb is too small to be cost competitive with high density NAND for mass storage (SSD’s smartphones etc) but Sony are surely targeting a different market segment. What that is a matter of speculation but a high performance cache for a SSD or traditional HD would be near the top of my list. If the performance gains over NAND in terms of speed and endurance are sufficiently compelling, embedding the chip in a larger system takes the pressure of the manufacturing cost of the chip.
3) Embedded ReRAM: 2013 saw the release of the Panasonic MCU with an embedded ReRAM memory. Here again the performance gains and reduction in power consumption drove the introduction of ‘new’ memory. I expect Panasonic to come up with a Mark 2 product and other companies to follow suit. But which companies?
4) Niche markets: Adesto have published results focussed on high temperature operation and radiation hardness suggesting products for specialized markets could be forthcoming in 2014.
5) Consolidation: At present we have the slightly odd situation where ReRAM/CBRAM development is being led by both small start-ups (Crossbar, Adesto, 4DS, Symetrix,…) all the major memory manufacturers (SK Hynix, Micron, Toshiba/SanDisk, Samsung) as well as large companies not currently in the memory manufacturing business (HP, IBM, Panasonic, Sony). I expected consolidation in 2013 along the lines of the 2012 purchase of Unity by Rambus. It did not take place and the start-ups look stronger (at least from a technology perspective) now than a year ago. So for 2014, I am going to punt on this one!
6) Neuromorphic computing: Lots of interest and academic research in 2013 which I expect to be followed by major breakthroughs in 2014. I’m still not convinced that an ReRAM cell can act as an ideal ‘neuron’ but I fully expect someone to come up with a way of working round this issue.
7) ReRAM-Forum will reach 100 posts! This is going to be tough and I can’t do it on my own. Please get in touch if you wish to post or have ideas for a post on the Blog. Other ways you can help is to let me know if you would like to be the focus of an interview or have a paper being published.
All the Best for a Happy and Successful 2014
Christie Marrian, ReRAM-Forum.com Moderator