Carbon nanotubes have been around for awhile and their characteristics have prompted researchers to try using them for storing bits. They can be very densely packed, so supporting huge storage capacity, they are fast and require very little energy. They seem an ideal replacement of flash memory.
The quest for an industrial process able to deliver a nanotube based storage chip (NRAM) started some 15 years ago but it proved elusive. The problem was to create an ordered structure for the nanotubes but that was too costly and the yield was very very low.
A start up, Nantero, proposed a different approach. Make use of randomly disposed nanotube and use them in such a way that the way of using was actually creating an ordered structure. The idea was interesting and worth pursuing.
Over these fifteen years over 70 million $ have been invested in trying to make this an industrial reality. In the process Nantero has been able to patent and sell its evolving technology to several other companies with an income approaching 50 million $.
Still, the manufacturing of a NRAM chip remained a dream. Besides, in these 15 years flash memory has continued to improve and NRAM seemed to be a lost bet.
Now, however, flash memory seems to have reached the end of the line in terms of evolution, whilst the need for storage capacity keeps growing. Hence the renovated interest for an alternative technology.
Nantero has secured at the end of 2016 an investment of 21 million $ and has announced they are now ready to deliver NRAM chip and they are partnering with Fujitsu to make the first nanotube based storage chip.
For the first time there is a date for go-to-market: 2018. They expect to develop the chip using 55nm technology and deliver a writing speed 1,000 times faster than today's flash memory with a life time (number of rewriting cycles) many thousands of times of today's flash memory.