Well... speak of the devil:
***
According to Professor Paul Heremans of the University of Leuven in Belgium, circuit designers looking for the best performance often have to combine several memory types on the same chip. This adds complexity and cost.
A more serious issue is scalability. As designers pack more components onto each chip, the width of the smallest features is shrinking, from 130 nanometers (nm) in 2000 to 45 nm today. Existing memory technologies are good for several more generations, Heremans says, but are unlikely to make the transition to 22 nm (scheduled for 2011) or 16 nm (2018).
So we need new memory technologies that can be made smaller than those of today, as well as preferably being faster, power saving and non-volatile. The runners in the global memory technology race form a veritable alphabet soup of acronyms including MRAM, RRAM, FeRAM, Z-RAM, SONOS, and Nano-RAM.
No universal solution
Early in 2004, Heremans became the coordinator of an EU-supported project that included two of Europe’s biggest semiconductor manufacturers: STMicroelectronics of Italy and Philips of the Netherlands. Heremans’ own institution, IMEC, is a leading independent research centre in microelectronics and nanotechnology. The Polish Academy of Sciences was the fourth partner in the project.
***
cordis.europa.eu
All this research activity will become even more frenetic as we get closer to 2012 and 22nm. Ultimately I don't think there will ever be a consensus on an ultimate winner. Nor do I believe this recurring theme of "No universal solution."
The costs involved would seem to guarantee a "universal solution" as far as the mainstream commercial markets are concerned. And rather than determining the "winner" through any abstract industry consensus on a new commercial memory, it seems more likely that the successful IP will arise through the same kind of fab process development and adoption that gave rise to today's Flash IP.
By that measure F-RAM is obviously in the lead when compared to the other candidates. SONOS is probably next in so far as volume production and applications are concerned, but I would expect that both MRAM and PCRAM/RRAM flavors will have to start showing up in significant real world products before 2009 if their IP is going to be a serious alternative by 2012.
I expect FSL/MOT's 1st generation MRAM to have less impact on end markets than F-RAM. So far I've only seen it on one FPGA development board and it will have to do a lot better next year to remain viable. TMSC, Toshiba, and IBM/IFX are all working on MRAM improvements for generations which could see commercial production in the 2009 - 2011 time frame. That's cutting it pretty close if Flash or CMOS development is going to "expire" in the 2012 - 2018 time frame.
But still I'd certainly put MRAM well ahead of PCRAM/RRAM. Not only is it in some form of commercial production, but it also seems to have more in common with 2 of the 3 potential technologies being considered as candidates to replace CMOS logic:
In addition, the new switch would be functionally scalable through multiple computer chip nodes. According to Kaloyeros, there are four candidates for non-charge based state variables. The switching is based on an exciton, a spin, orbital symmetry and magnetic flux quantum.
eetimes.com
See also;
en.wikipedia.org
So the question is... assuming the potential CMOS replacements can really be limited to the above (which is questionable in my view)... will L'Intel/STM really pull the trigger on PCRAM/RRAM if they're going to have to compete with FLASH IP pricing through 2012 or 2018... and then have to deal with an incompatible CMOS IP almost as soon as Flash gives up the ghost? I certainly wouldn't do it... but then L'Intel has been known to dump tons of cash into non-commercial propositions before and their NOR itself has been a drain on income for sometime. Is there a high volume application out there for 128Mb PCRAM that simply can't use anything else? We shall see.
0|0 |
