Lowrey gave a nice primer and update on OUM.
Measures of success:
Expansion of JV with STM and Intel is a very significant endorsement of the viability of OUM for semi memory applications.
Working memory in competitive sized arrays at all 3 licensees.
'Market is moving rapidly to NV memory'
Low voltage and low power -- 'we have a very big advantage there' as per Lowrey
OUM as a top candidate for embedded memory in uProcessors, ASP's, DSP's, etc. These applications move OUM up from the jelly-bean dedicated memory arena.
Extrinsic (outlier) faulty bits still the prime focus of failure mode analysis. Always will be for OUM as well as for any, and all, memory products. Indications are they're fairly advanced along the learning curve.
0.18um technology used for test chips -- hinted that work being done at 0.13um -- another favorable outlook for the future of OUM - as they, and I, perceive it.
Data retention slide, before and after a 140C 16 hr bake, shows a higher resistance after the bake -- opposite to what I initially thought would be occurring. The previous graph shows a faster movement from amorphous-to-crystalline with higher temp. exposures, a sort of slow annealing process. This would lower the resistance with time. Somewhat intrigued with the minimum resistance level shown on the graph (approx 100,000 ohms). Freely interpreting the graph, I expect that the 2 orders of magnitude (10^5-10^7) resistance range covers the extremes of crystalline-to-amorphous so the structure IMO is possibly a high aspect ratio via (hole) with a small crystalline volume adjacent to the heating electrode and a relative extensive length of amorphous material above that to the upper contact. Perhaps other mechanisms are operative with this structure. Any thoughts RAY?
BAE is very successful in adapting OUM for NV space environments. Practically no competition re OUM's robustness to radiation damage. Too bad some of the space programs are being stalled budgetwise - re the EUROPA program. A while back on the YAHOO board, a person , 'where-is-fred', close to the procurement and product specification areas for NV memory, lamented the inadequacies of NV memory for space programs and he ventured that the 1st product that overcomes present limitations will be a huge winner, not only because of price premiums, but also that the volume need is large.
Overall I am well satisfied with the progress of this technology. It's prime competitor at present is multi-bit flash but as Lowrey pointed out the inherent limitations of FLASH require a radically new memory technology -- with OUM the prime candidate.
Ed |
