Tag Archives: computer memory

Memory material with functions resembling synapses and neurons in the brain

Leave a reply

This work comes from the University of Twente’s MESA+ Institute for Nanotechnology according to a July 8, 2016 news item on ScienceDaily,

Our brain does not work like a typical computer memory storing just ones and zeroes: thanks to a much larger variation in memory states, it can calculate faster consuming less energy. Scientists of the MESA+ Institute for Nanotechnology of the University of Twente (The Netherlands) now developed a ferro-electric material with a memory function resembling synapses and neurons in the brain, resulting in a multistate memory. …

A July 8, 2016 University of Twente press release, which originated the news item, provides more technical detail,

The material that could be the basic building block for ‘brain-inspired computing’ is lead-zirconium-titanate (PZT): a sandwich of materials with several attractive properties. One of them is that it is ferro-electric: you can switch it to a desired state, this state remains stable after the electric field is gone. This is called polarization: it leads to a fast memory function that is non-volatile. Combined with processor chips, a computer could be designed that starts much faster, for example. The UT scientists now added a thin layer of zinc oxide to the PZT, 25 nanometer thickness. They discovered that switching from one state to another not only happens from ‘zero’ to ‘one’ vice versa. It is possible to control smaller areas within the crystal: will they be polarized (‘flip’) or not?

In a PZT layer without zinc oxide (ZnO) there are basically two memorystates. Adding a nano layer of ZnO, every state in between is possible as well.

Multistate

By using variable writing times in those smaller areas, the result is that many states can be stored anywhere between zero and one. This resembles the way synapses and neurons ‘weigh’ signals in our brain. Multistate memories, coupled to transistors, could drastically improve the speed of pattern recognition, for example: our brain performs this kind of tasks consuming only a fraction of the energy a computer system needs. Looking at the graphs, the writing times seem quite long compared to nowaday’s processor speeds, but it is possible to create many memories in parallel. The function of the brain has already been mimicked in software like neurale networks, but in that case conventional digital hardware is still a limitation. The new material is a first step towards electronic hardware with a brain-like memory. Finding solutions for combining PZT with semiconductors, or even developing new kinds of semiconductors for this, is one of the next steps.

Here’s a link to and a citation for the paper,

Multistability in Bistable Ferroelectric Materials toward Adaptive Applications by Anirban Ghosh, Gertjan Koster, and Guus Rijnders. Advanced Functional Materials DOI: 10.1002/adfm.201601353 Version of Record online: 4 JUL 2016

© 2016 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

This paper is behind a paywall.

Nucleic acid-based memory storage

Leave a reply

We’re running out of memory. To be more specific, there are two problems: the supply of silicon and a limit to how much silicon-based memory can store. An April 27, 2016 news item on Nanowerk announces a nucleic acid-based approach to solving the memory problem,

A group of Boise State [Boise State University in Idaho, US] researchers, led by associate professor of materials science and engineering and associate dean of the College of Innovation and Design Will Hughes, is working toward a better way to store digital information using nucleic acid memory (NAM).

An April 25, 2016 Boise State University news release, which originated the news item, expands on the theme of computer memory and provides more details about the approach,

It’s no secret that as a society we generate vast amounts of data each year. So much so that the 30 billion watts of electricity used annually by server farms today is roughly equivalent to the output of 30 nuclear power plants.

And the demand keeps growing. The global flash memory market is predicted to reach $30.2 billion this year, potentially growing to $80.3 billion by 2025. Experts estimate that by 2040, the demand for global memory will exceed the projected supply of silicon (the raw material used to store flash memory). Furthermore, electronic memory is rapidly approaching its fundamental size limits because of the difficulty in storing electrons in small dimensions.

Hughes, with post-doctoral researcher Reza Zadegan and colleagues Victor Zhirnov (Semiconductor Research Corporation), Gurtej Sandhun (Micron Technology Inc.) and George Church (Harvard University), is looking to DNA molecules to solve the problem. Nucleic acid — the “NA” in “DNA” — far surpasses electronic memory in retention time, according to the researchers, while also providing greater information density and energy of operation.

Their conclusions are outlined in an invited commentary in the prestigious journal Nature Materials published earlier this month.

“DNA is the data storage material of life in general,” said Hughes. “Because of its physical and chemical properties, it also may become the data storage material of our lives.” It may sound like science fiction, but Hughes will participate in an invitation-only workshop this month at the Intelligence Advanced Research Projects Activity (IARPA) Agency to envision a portable DNA hard drive that would have 500 Terabytes of searchable data – that’s about the the size of the Library of Congress Web Archive.

“When information bits are encoded into polymer strings, researchers and manufacturers can manage and manipulate physical, chemical and biological information with standard molecular biology techniques,” the paper [in Nature Materials?] states.

Cost-competitive technologies to read and write DNA could lead to real-world applications ranging from artificial chromosomes, digital hard drives and information-management systems, to a platform for watermarking and tracking genetic content or next-generation encryption tools that necessitate physical rather than electronic embodiment.

Here’s how it works. Current binary code uses 0’s and 1’s to represent bits of information. A computer program then accesses a specific decoder to turn the numbers back into usable data. With nucleic acid memory, 0’s and 1’s are replaced with the nucleotides A, T, C and G. Known as monomers, they are covalently bonded to form longer polymer chains, also known as information strings.

Because of DNA’s superior ability to store data, DNA can contain all the information in the world in a small box measuring 10 x 10 x 10 centimeters cubed. NAM could thus be used as a sustainable time capsule for massive, scientific, financial, governmental, historical, genealogical, personal and genetic records.

Better yet, DNA can store digital information for a very long time – thousands to millions of years. Currently, usable information has been extracted from DNA in bones that are 700,000 years old, making nucleic acid memory a promising archival material. And nucleic acid memory uses 100 million times less energy than storing data electronically in flash, and the data can live on for generations.

At Boise State, Hughes and Zadegan are examining DNA’s stability under extreme conditions. DNA strands are subjected to temperatures varying from negative 20 degrees Celsius to 100 degrees Celsius, and to a variety of UV exposures to see if they can still retain their information. What they’re finding is that much less information is lost with NAM than with the current state of the industry.

Here’s a link to and a citation for the Nature Materials paper,

Nucleic acid memory by Victor Zhirnov, Reza M. Zadegan, Gurtej S. Sandhu, George M. Church, & William L. Hughes. Nature Materials 15, 366–370 (2016)  doi:10.1038/nmat4594 Published online 23 March 2016

This paper is behind a paywall.

Memristor shakeup

Leave a reply

New discoveries suggest that memristors do not function as was previously theorized. (For anyone who wants a memristor description, there’s this Wikipedia entry.) From an Oct. 13, 2015 posting by Alexander Hellemans for the Nanoclast blog (on the IEEE [Institute for Electrical and Electronics Engineers]), Note: Links have been removed,

What’s going to replace flash? The R&D arms of several companies including Hewlett Packard, Intel, and Samsung think the answer might be memristors (also called resistive RAM, ReRAM, or RRAM). These devices have a chance at unseating the non-volatile memory champion because, they use little energy, are very fast, and retain data without requiring power. However, new research indicates that they don’t work in quite the way we thought they do.

The fundamental mechanism at the heart of how a memristor works is something called an “imperfect point contact,” which was predicted in 1971, long before anybody had built working devices. When voltage is applied to a memristor cell, it reduces the resistance across the device. This change in resistance can be read out by applying another, smaller voltage. By inverting the voltage, the resistance of the device is returned to its initial value, that is, the stored information is erased.

Over the last decade researchers have produced two commercially promising types of memristors: electrochemical metallization memory (ECM) cells, and valence change mechanism memory (VCM) cells.

…

Now international research teams lead by Ilia Valov at the Peter Grünberg Institute in Jülich, Germany, report in Nature Nanotechnology and Advanced Materials that they have identified new processes that erase many of the differences between EMC and VCM cells.

Valov and coworkers in Germany, Japan, Korea, Greece, and the United States started investigating memristors that had a tantalum oxide electrolyte and an active tantalum electrode. “Our studies show that these two types of switching mechanisms in fact can be bridged, and we don’t have a purely oxygen type of switching as was believed, but that also positive [metal] ions, originating from the active electrode, are mobile,” explains Valov.

Here are links to and citations for both papers,

Graphene-Modified Interface Controls Transition from VCM to ECM Switching Modes in Ta/TaOx Based Memristive Devices by Michael Lübben, Panagiotis Karakolis, Vassilios Ioannou-Sougleridis, Pascal Normand, Pangiotis Dimitrakis, & Ilia Valov. Advanced Materials DOI: 10.1002/adma.201502574 First published: 10 September 2015

Nanoscale cation motion in TaOx, HfOx and TiOx memristive systems by Anja Wedig, Michael Luebben, Deok-Yong Cho, Marco Moors, Katharina Skaja, Vikas Rana, Tsuyoshi Hasegawa, Kiran K. Adepalli, Bilge Yildiz, Rainer Waser, & Ilia Valov. Nature Nanotechnology (2015) doi:10.1038/nnano.2015.221 Published online 28 September 2015

Both papers are behind paywalls.