Science

New "ReRAM" memory chip outdoes flash memory in speed, density and energy efficiency

View 4 Images
UCL researchers have designed a silicon oxide-based memory chip that is faster than flash memory (Photo: UCL/Adnan Mehonic)
UCL researchers have designed a silicon oxide-based memory chip that is faster than flash memory (Photo: UCL/Adnan Mehonic)
The new chip is cheap to manufacture because it does not require a vacuum to work and can operate in ambient conditions (Image: UCL/Adnan Mehonic)
The chip is made with oxides of metals that can remember the change in electrical resistance when they receive a discharge of electricity (Image: UCL/Adnan Mehonic)
UCL researchers have designed a silicon oxide-based memory chip that is faster than flash memory (Photo: UCL/Adnan Mehonic)
View gallery - 4 images

These days, Flash memory is almost the defacto standard for data storage in consumer devices, being found in everything from PCs and digital cameras, to smartphones and USB thumb drives. But a team of researchers at University College of London (UCL) has developed a new type of memory chip that is much faster than Flash memory, while also offering greater storage densities and requiring much less power. Could the days of Flash memory's dominance be numbered?

The new chip developed at UCL is the first purely silicon oxide-based type of memory chip known as Resistive RAM, (ReRAM). These are memory chips based on materials, most commonly oxides of metals, that can remember the change in electrical resistance when a voltage is applied. The chip is also a so-called “memristor," a kind of electronic holy grail that took decades to be prototyped, with HP researchers finally clearing that hurdle in 2008 with the development of the first practical memristor that was based on titanium dioxide.

Memristors work similarly to brain neurons, with a continuously variable resistance that can be changed according to the current that passes through them and the value remembered once the power is turned off. This property, along with the fact the UCL team's new chip is based on silicon oxide, opens up the possibility of numerous applications, including their potential incorporation into not only memory, but also computer processors.

Because it does not require a vacuum to work and can operate in ambient conditions, the team says its chip is potentially cheaper and more durable than other silicon oxide chips currently in development. The chip's design also opens up the possibility of transparent memory chips for use in touch screens and mobile devices, with the team already working on making a quartz device that could be used in such transparent electronics.

Several top semiconductor companies have already approached the UCL researchers, attracted by the chip’s game-changing potential. “Our ReRAM memory chips need just a thousandth of the energy and are around a hundred times faster than standard Flash memory chips,” said Dr Tony Kenyon from UCL's Electronic and Electrical Engineering department.

Like a lot of great scientific discoveries, ReRAM was discovered by accident. The researchers were handling silicon oxide material to produce silicon LEDs when they noticed the devices appeared unstable. PhD student Adnan Mehonic looked into it and had his own eureka moment when he discovered that in reality the material alternated between conducting and non-conducting states in a predictable pattern that revealed memristor potential.

The structure and performance of the silicon oxide structure (and its switch in resistance) was recently described in a paper published in the Journal of Applied Physics.

Source: UCL

View gallery - 4 images
  • Facebook
  • Twitter
  • Flipboard
  • LinkedIn
1 comment
Lazlo
There's no mention of manufacturing cost / scalability. How about requiring major new manufacturing processes or can most of their current equipment be re-purposed? Are they going to retain tight control or are they planning to license relatively cheaply to anyone that wants it? Some things in the article indicate that these may not be issues, such as interest from manufacturers, but it's a fairly important issue when the question of your article is could it replace the current standard. Adoption of new tech usually has a lot more to do with economics than technical specs.
I am a sucker for "Eureka!" moments. That this was noticed and recognized for what it was by PhD student is even better. Imho they should award him his doctorate based on this. I wouldn't be surprised if I read his name again in the future.