Whatever happened to …

[Grumble]

Memristors?

Ten years ago or so they were the next big thing, a new fundamental building block of electronic circuits. Not really heard of them since. Shouldn’t they be in all our phones by now?

[shpalman]

We went with flash memory instead.

[Grumble]

We went with flash memory instead.

Flash memory well predates memristors though, which were theoretical devices predicted in the 1960’s or possibly earlier, only becoming real when silicon fabs got small scale enough for quantum effects to become significant in the 2010’s. At least that’s my memory.

[Grumble]

https://en.m.wikipedia.org/wiki/Memristor

[Martin Y]

I think memristors passed me by. When I was a student forty years ago a couple of people in the electronics dept were doing work on Bubble Memory which was going to be the next big thing.

[IvanV]

https://en.m.wikipedia.org/wiki/Memristor

I think that is rather out of date. It says there,

Memristors remain a laboratory curiosity, as yet made in insufficient numbers to gain any commercial applications. Despite this lack of mass availability, according to Allied Market Research [writing in 2017] the memristor market was worth $3.2 million in 2015 and will be worth $79.0 million by 2022.

But it seems that situation is being rapidly being overturned, as numerous companies have gone into manufacture and the memristor market is growing faster than that forecast.

The memristor market was already worth $190m in 2022. High speeds of growth are forecast in the next few years. Though of course forecasts are often wrong, the above one was, though at least they got the general idea of rapid growth correct.

As IMARC say in that recent market study I pointed to,

Due to the key advantages of memristors, including compact size, high retention, enhanced endurance, three-dimensional (3D) design capabilities, and fast reading and writing speed, they have emerged as highly reliable alternatives to flash memories. This, in confluence with the escalating demand for portable, miniature electronics, is influencing the market positively. Moreover, the emerging automation trend in industrial processes is promoting the adoption of robots, which, in turn, is driving the market. Apart from this, the increasing utilization of data centers in the healthcare, information and technology (IT) and telecommunications, and banking, financial services, and insurance (BFSI) industries is impelling the growth of the market worldwide. In line with this, governments of several countries are investing in improving IT infrastructure, which is creating a favorable outlook for the market. Besides this, the increasing investment in research activities to develop innovative transparent electronics, which rely on memristors for storage purposes, is projected to strengthen the market growth in the coming years. Transparent electronics find extensive applications in transportation, energy sources, consumer electronics, and automobile windshields that can transmit visual information to the driver.

[Boustrophedon]

As IMARC say in that recent market study I pointed to,

Due to the key advantages of memristors, including compact size, high retention, enhanced endurance, three-dimensional (3D) design capabilities, and fast reading and writing speed, they have emerged as highly reliable alternatives to flash memories. This, in confluence with the escalating demand for portable, miniature electronics, is influencing the market positively. Moreover, the emerging automation trend in industrial processes is promoting the adoption of robots, which, in turn, is driving the market. Apart from this, the increasing utilization of data centers in the healthcare, information and technology (IT) and telecommunications, and banking, financial services, and insurance (BFSI) industries is impelling the growth of the market worldwide. In line with this, governments of several countries are investing in improving IT infrastructure, which is creating a favorable outlook for the market. Besides this, the increasing investment in research activities to develop innovative transparent electronics, which rely on memristors for storage purposes, is projected to strengthen the market growth in the coming years. Transparent electronics find extensive applications in transportation, energy sources, consumer electronics, and automobile windshields that can transmit visual information to the driver.

That reads very oddly, almost as if it were generated by AI.

[shpalman]

The flash memory market is worth more like $80 billion.

That reads like something motivating why ICT will keep growing in general, not that there's any specific issue with flash memory which memristors will solve. They may always remain about 1% of the flash memory market value.

[shpalman]

I was just reminded to check whatever happened to those CNTs at Nantero and what they're claiming they'll be doing in two years' time this time.*

Over the past twenty years we seem to have gone from memory cells based on individual nanotubes touching and then separating** to some sort of sh.t which you dump all over a wafer and then find conductive paths through it.

* For example in 2016 they claimed this for 2018 https://www.tomshardware.com/news/fujit ... 32603.html

** An idea which is obviously impossible, as anyone who has considered the basic physics for a moment would have figured out (not even considering the total impracticality of placing individual nanotubes with the right properties into the device in the right places). In fact it never worked, what actually happened was that there'd be a whole bunch of nanotubes stuck together by Van der Waals forces and by passing a current through it they'd "solder" themselves together and then by passing a higher current you could burn them out. So it would switch once but then not work anymore.