Toward brain-like computing: New memristor better mimics synapses
A new electronic device developed at the University of Michigan (U-M) can directly model the behaviors of a synapse, which is a connection between two neurons.
For the first time, the way that neurons share or compete for resources can be explored in hardware without the need for complicated circuits.
"Neuroscientists have argued that competition and cooperation behaviors among synapses are very important. Our new memristive devices allow us to implement a faithful model of these behaviors in a solid-state system," said Wei Lu, U-M professor of electrical and computer engineering and senior author of the study in Nature Materials.
[...] The memristor is a good model for a synapse. It mimics the way that the connections between neurons strengthen or weaken when signals pass through them. But the changes in conductance typically come from changes in the shape of the channels of conductive material within the memristor. These channels—and the memristor's ability to conduct electricity—could not be precisely controlled in previous devices.
Now, the U-M team has made a memristor in which they have better command of the conducting pathways.They developed a new material out of the semiconductor molybdenum disulfide—a "two-dimensional" material that can be peeled into layers just a few atoms thick. Lu's team injected lithium ions into the gaps between molybdenum disulfide layers.
They found that if there are enough lithium ions present, the molybdenum sulfide transforms its lattice structure, enabling electrons to run through the film easily as if it were a metal. But in areas with too few lithium ions, the molybdenum sulfide restores its original lattice structure and becomes a semiconductor, and electrical signals have a hard time getting through.
Related: This Tiny Electronic Chip Is Just 3 Atoms Thick
A New Generation of Artificial Retinas Based on 2-D Materials
Purdue University Researchers Identify Molybdenum Ditelluride as a Material for Next-Gen Memory
Ionic modulation and ionic coupling effects in MoS2 devices for neuromorphic computing (DOI: 10.1038/s41563-018-0248-5) (DX)
(Score: 2) by hendrikboom on Tuesday December 18 2018, @11:51PM (7 children)
What *is* a memristor? What electrical properties make it a memory device?
(Score: 2, Informative) by NPC-131072 on Wednesday December 19 2018, @01:19AM (5 children)
All answered here [arstechnica.com]
(Score: 2) by takyon on Wednesday December 19 2018, @01:06PM (3 children)
I have the sneaking suspicion you're building your account's hit points so you can take a few hits. :-)
[SIG] 10/28/2017: Soylent Upgrade v14 [soylentnews.org]
(Score: 0) by Anonymous Coward on Wednesday December 19 2018, @07:26PM (1 child)
That would be a cool idea for a social media site. Karma > XP, every time you max out karma, get a new life (level), get demoted to level -1, no posting for a week.
(Score: 2) by takyon on Wednesday December 19 2018, @07:34PM
If karma was not capped at 50 here, I'd probably have well over 20,000. I could then use that accumulated karma to advance up and become a Lvl. 69 White Knight.
[SIG] 10/28/2017: Soylent Upgrade v14 [soylentnews.org]
(Score: 1) by NPC-131072 on Wednesday December 19 2018, @10:57PM
Hello fren, it's not easy being the vanguard against nazis and transmisogynists. Also, "electrical properties" is not straightforward as there are different implementations.
(Score: 2) by hendrikboom on Wednesday December 19 2018, @03:13PM
Thanks. That was helpful.
(Score: 0) by Anonymous Coward on Wednesday December 19 2018, @07:40PM
Is this 'injection' of lithium possible with mass manufacturing fabrication techniques, or only in a lab? And if it is possible in the former case, does this process actually function through entirely electrical means? If it cannot be controlled easily by electricity, or has a sufficiently low read/write life for the cells, then it won't matter if it works if it is effectively single use (not unlike a brain in fact.)
A followup question is if the current resistances of a particular memristor could be read by a conventional semiconduction circuit manufactured on the same substrate alongside it, and then be used to back up the state of each individual memristor in the design, or additionally be used to prep a particular memristor to a desired value without having to initialize all memristors with the same initial stimulus used to create the original memristor 'brain'.
Answer these questions and it might finally be possible to not only emulate a human-like brain using semiconductor processors, but also back up and restore it during the development and testing of such a design. Until all of these facets are in place, any produced synthetic brains would be just as useless as a human one for creating a reproducible artificial consciousness.