The ISSCC Advance Program is now public and I am continuing to see a disturbing trend there and at IEDM with respect to emerging memory technologies. I first wrote about this apparent trend late last year and I wanted to give an update.

IMHO there is a significant disparity between the opportunity presented for emerging memory by the difficulty of continued DRAM scaling and of increasing NAND layer counts (perhaps less so) versus the apparent industry activity based on the frequency of certain reference terms in industry conference papers. I wanted to try and quantify this so I created the chart below – an aggregate of the number of times certain emerging memory key terms (“RRAM”/”ReRAM”, “MRAM”, etc.) appear in the IEDM and ISSCC programs – IEDM and ISSCC being leading device and circuit technical conferences.

As one can see, there was a burst of activity 6-7 years ago (perhaps prompted by Micron/Intel 3D-xpoint interest as well as a wave of MRAM advancements); however, the apparent activity has fallen off year over year since then.

I’m not going to try and interpret the overall meaning of this trend; however, I do find it disappointing to not see more emerging memory technology papers in these conferences, and ultimately a successful productization (non-embedded) of one of these technologies!

That said I am seeing what I believe is increasing activity regarding emerging memory (RRAM in particular) cells used in compute-in-memory/neuromorphic applications, and I suspect we will see much more of this to come.