Engineers put tens of thousands of artificial brain synapses on a single chip

Brain on chip abstract concept

MIT engineers have designed a "brain-on-a-chip," smaller than a piece of confetti, that is made from tens of thousands of artificial brain synapses known as memristors -- silicon-based components that mimic the information-transmitting synapses in the human brain.

The researchers borrowed from principles of metallurgy to fabricate each memristor from alloys of silver and copper, along with silicon. When they ran the chip through several visual tasks, the chip was able to "remember" stored images and reproduce them many times over, in versions that were crisper and cleaner compared with existing memristor designs made with unalloyed elements.

Their results, published today in the journal Nature Nanotechnology, demonstrate a promising new memristor design for neuromorphic devices -- electronics that are based on a new type of circuit that processes information in a way that mimics the brain's neural architecture. Such brain-inspired circuits could be built into small, portable devices, and would carry out complex computational tasks that only today's supercomputers can handle.

"So far, artificial synapse networks exist as software. We're trying to build real neural network hardware for portable artificial intelligence systems," says Jeehwan Kim, associate professor of mechanical engineering at MIT. "Imagine connecting a neuromorphic device to a camera on your car, and having it recognize lights and objects and make a decision immediately, without having to connect to the internet. We hope to use energy-efficient memristors to do those tasks on-site, in real-time."

Wandering ions

Memristors, or memory transistors, are an essential element in neuromorphic computing. In a neuromorphic device, a memristor would serve as the transistor in a circuit, though its workings would more closely resemble a brain synapse -- the junction between two neurons. The synapse receives signals from one neuron, in the form of ions, and sends a corresponding signal to the next neuron.

A transistor in a conventional circuit transmits information by switching between one of only two values, 0 and 1, and doing so only when the signal it receives, in the form of an electric current, is of a particular strength. In contrast, a memristor would work along a gradient, much like a synapse in the brain. The signal it produces would vary depending on the strength of the signal that it receives. This would enable a single memristor to have many values, and therefore carry out a far wider range of operations than binary transistors.

Like a brain synapse, a memristor would also be able to "remember" the value associated with a given current strength, and produce the exact same signal the next time it receives a similar current. This could ensure that the answer to a complex equation, or the visual classification of an object, is reliable -- a feat that normally involves multiple transistors and capacitors.

Ultimately, scientists envision that memristors would require far less chip real estate than conventional transistors, enabling powerful, portable computing devices that do not rely on supercomputers, or even connections to the Internet.

Existing memristor designs, however, are limited in their performance. A single memristor is made of a positive and negative electrode, separated by a "switching medium," or space between the electrodes. When a voltage is applied to one electrode, ions from that electrode flow through the medium, forming a "conduction channel" to the other electrode. The received ions make up the electrical signal that the memristor transmits through the circuit. The size of the ion channel (and the signal that the memristor ultimately produces) should be proportional to the strength of the stimulating voltage.

Kim says that existing memristor designs work pretty well in cases where voltage stimulates a large conduction channel, or a heavy flow of ions from one electrode to the other. But these designs are less reliable when memristors need to generate subtler signals, via thinner conduction channels.

The thinner a conduction channel, and the lighter the flow of ions from one electrode to the other, the harder it is for individual ions to stay together. Instead, they tend to wander from the group, disbanding within the medium. As a result, it's difficult for the receiving electrode to reliably capture the same number of ions, and therefore transmit the same signal, when stimulated with a certain low range of current.

Borrowing from metallurgy

Kim and his colleagues found a way around this limitation by borrowing a technique from metallurgy, the science of melding metals into alloys and studying their combined properties.

"Traditionally, metallurgists try to add different atoms into a bulk matrix to strengthen materials, and we thought, why not tweak the atomic interactions in our memristor, and add some alloying element to control the movement of ions in our medium," Kim says.

Engineers typically use silver as the material for a memristor's positive electrode. Kim's team looked through the literature to find an element that they could combine with silver to effectively hold silver ions together, while allowing them to flow quickly through to the other electrode.

The team landed on copper as the ideal alloying element, as it is able to bind both with silver, and with silicon.

"It acts as a sort of bridge, and stabilizes the silver-silicon interface," Kim says.

To make memristors using their new alloy, the group first fabricated a negative electrode out of silicon, then made a positive electrode by depositing a slight amount of copper, followed by a layer of silver. They sandwiched the two electrodes around an amorphous silicon medium. In this way, they patterned a millimeter-square silicon chip with tens of thousands of memristors.

As a first test of the chip, they recreated a gray-scale image of the Captain America shield. They equated each pixel in the image to a corresponding memristor in the chip. They then modulated the conductance of each memristor that was relative in strength to the color in the corresponding pixel.

The chip produced the same crisp image of the shield, and was able to "remember" the image and reproduce it many times, compared with chips made of other materials.

The team also ran the chip through an image processing task, programming the memristors to alter an image, in this case of MIT's Killian Court, in several specific ways, including sharpening and blurring the original image. Again, their design produced the reprogrammed images more reliably than existing memristor designs.

"We're using artificial synapses to do real inference tests," Kim says. "We would like to develop this technology further to have larger-scale arrays to do image recognition tasks. And some day, you might be able to carry around artificial brains to do these kinds of tasks, without connecting to supercomputers, the internet, or the cloud."

New light shed on intelligent life existing across the galaxy

Illustration of alien world, starry background

One of the biggest and longest-standing questions in the history of human thought is whether there are other intelligent life forms within our Universe. Obtaining good estimates of the number of possible extraterrestrial civilizations has however been very challenging.

A new study led by the University of Nottingham and published today in The Astrophysical Journal has taken a new approach to this problem. Using the assumption that intelligent life forms on other planets in a similar way as it does on Earth, researchers have obtained an estimate for the number of intelligent communicating civilizations within our own galaxy -- the Milky Way. They calculate that there could be over 30 active communicating intelligent civilizations in our home Galaxy.

Professor of Astrophysics at the University of Nottingham, Christopher Conselice who led the research, explains: "There should be at least a few dozen active civilizations in our Galaxy under the assumption that it takes 5 billion years for intelligent life to form on other planets, as on Earth." Conselice also explains that, "The idea is looking at evolution, but on a cosmic scale. We call this calculation the Astrobiological Copernican Limit."

First author Tom Westby explains: "The classic method for estimating the number of intelligent civilizations relies on making guesses of values relating to life, whereby opinions about such matters vary quite substantially. Our new study simplifies these assumptions using new data, giving us a solid estimate of the number of civilizations in our Galaxy.

The two Astrobiological Copernican limits are that intelligent life forms in less than 5 billion years, or after about 5 billion years -- similar to on Earth where a communicating civilization formed after 4.5 billion years. In the strong criteria, whereby a metal content equal to that of the Sun is needed (the Sun is relatively speaking quite metal rich), we calculate that there should be around 36 active civilizations in our Galaxy."

The research shows that the number of civilizations depends strongly on how long they are actively sending out signals of their existence into space, such as radio transmissions from satellites, television, etc. If other technological civilizations last as long as ours which is currently 100 years old, then there will be about 36 ongoing intelligent technical civilizations throughout our Galaxy.

However, the average distance to these civilizations would be 17,000 light-years away, making detection and communication very difficult with our present technology. It is also possible that we are the only civilization within our Galaxy unless the survival times of civilizations like our own are long.

Professor Conselice continues: "Our new research suggests that searches for extraterrestrial intelligent civilizations not only reveals the existence of how life forms, but also gives us clues for how long our own civilization will last. If we find that intelligent life is common then this would reveal that our civilization could exist for much longer than a few hundred years, alternatively if we find that there are no active civilizations in our Galaxy it is a bad sign for our own long-term existence. By searching for extraterrestrial intelligent life -- even if we find nothing -- we are discovering our own future and fate."

As many as six billion Earth-like planets in our galaxy, according to new estimates

Exoplanet illustration

There may be as many as one Earth-like planet for every five Sun-like stars in the Milky way Galaxy, according to new estimates by University of British Columbia astronomers using data from NASA's Kepler mission.

To be considered Earth-like, a planet must be rocky, roughly Earth-sized and orbiting Sun-like (G-type) stars. It also has to orbit in the habitable zones of its star -- the range of distances from a star in which a rocky planet could host liquid water, and potentially life, on its surface.

"My calculations place an upper limit of 0.18 Earth-like planets per G-type star," says UBC researcher Michelle Kunimoto, co-author of the new study in The Astronomical Journal. "Estimating how common different kinds of planets are around different stars can provide important constraints on planet formation and evolution theories, and help optimize future missions dedicated to finding exoplanets."

According to UBC astronomer Jaymie Matthews: "Our Milky Way has as many as 400 billion stars, with seven per cent of them being G-type. That means less than six billion stars may have Earth-like planets in our Galaxy."

Previous estimates of the frequency of Earth-like planets range from roughly 0.02 potentially habitable planets per Sun-like star, to more than one per Sun-like star.

Typically, planets like Earth are more likely to be missed by a planet search than other types, as they are so small and orbit so far from their stars. That means that a planet catalogue represents only a small subset of the planets that are actually in orbit around the stars searched. Kunimoto used a technique known as 'forward modelling' to overcome these challenges.

"I started by simulating the full population of exoplanets around the stars Kepler searched," she explained. "I marked each planet as 'detected' or 'missed' depending on how likely it was my planet search algorithm would have found them. Then, I compared the detected planets to my actual catalogue of planets. If the simulation produced a close match, then the initial population was likely a good representation of the actual population of planets orbiting those stars."

Kunimoto's research also shed more light on one of the most outstanding questions in exoplanet science today: the 'radius gap' of planets. The radius gap demonstrates that it is uncommon for planets with orbital periods less than 100 days to have a size between 1.5 and two times that of Earth. She found that the radius gap exists over a much narrower range of orbital periods than previously thought. Her observational results can provide constraints on planet evolution models that explain the radius gap's characteristics.

Up to 45 percent of SARS-CoV-2 infections may be asymptomatic

Coronavirus in crowd,

An extraordinary percentage of people infected by the virus behind the ongoing deadly COVID-19 pandemic never show symptoms of the disease, according to the results of a Scripps Research analysis of public datasets on asymptomatic infections.

The findings, published in Annals of Internal Medicine, suggest that asymptomatic infections may account for as much as 45 percent of all COVID-19 cases, playing a significant role in the early and ongoing spread of COVID-19. The report highlights the need for expansive testing and contact tracing to mitigate the pandemic.

"The silent spread of the virus makes it all the more challenging to control," says Eric Topol, MD, founder and director of the Scripps Research Translational Institute and professor of Molecular Medicine at Scripps Research. "Our review really highlights the importance of testing. It's clear that with such a high asymptomatic rate, we need to cast a very wide net, otherwise the virus will continue to evade us."

Together with behavioral scientist Daniel Oran, Topol collected information from testing studies on 16 diverse cohorts from around the world. These datasets -- gathered via keyword searches of PubMed, bioRxiv and medRxiv, as well as Google searches of relevant news reports -- included data on nursing home residents, cruise ship passengers, prison inmates and various other groups.

"What virtually all of them had in common was that a very large proportion of infected individuals had no symptoms," says Oran. "Among more than 3,000 prison inmates in four states who tested positive for the coronavirus, the figure was astronomical: 96 percent asymptomatic."

The review further suggests that asymptomatic individuals are able to transmit the virus for an extended period of time, perhaps longer than 14 days. The viral loads are very similar in people with or without symptoms, but it remains unclear whether their infectiousness is of the same magnitude. To resolve that issue, we'll need large-scale studies that include sufficient numbers of asymptomatic people.

The authors also conclude that the absence of symptoms may not imply an absence of harm. CT scans conducted on 54 percent of 76 asymptomatic individuals on the Diamond Princess cruise ship, appear to show significant subclinical lung abnormalities raising the possibility of SARS-CoV-2 infection impacting lung function that might not be immediately apparent. The scientists say further research is needed to confirm the potential significance of this finding.

The authors also acknowledge that the lack of longitudinal data makes distinguishing between asymptomatic and presymptomatic individuals difficult. An asymptomatic individual is someone who is infected with SARS-CoV-2, but never develops symptoms of COVID-19, while a presymptomatic person is similarly infected, but will eventually develop symptoms. Longitudinal testing, which refers to repeated testing of individuals over time, would help differentiate between the two.

"Our estimate of 40 to 45 percent asymptomatic means that, if you're unlucky enough to get infected, the probability is almost a flip of a coin on whether you're going to have symptoms. So to protect others, we think that wearing a mask makes a lot of sense," Oran concludes.

Face masks critical in preventing spread of COVID-19

Crowd of people wearing medical masks

A study by a team of researchers led by a Texas A&M University professor has found that not wearing a face mask dramatically increases a person's chances of being infected by the COVID-19 virus.

Renyi Zhang, Texas A&M Distinguished Professor of Atmospheric Sciences and the Harold J. Haynes Chair in the College of Geosciences, and colleagues from the University of Texas, the University of California-San Diego and the California Institute of Technology have had their work published in the current issue of PNAS (Proceedings of the National Academy of Sciences).

The team examined the chances of COVID-19 infection and how the virus is easily passed from person to person. From trends and mitigation procedures in China, Italy and New York City, the researchers found that using a face mask reduced the number of infections by more than 78,000 in Italy from April 6-May 9 and by over 66,000 in New York City from April 17-May 9.

"Our results clearly show that airborne transmission via respiratory aerosols represents the dominant route for the spread of COVID-19," Zhang said. "By analyzing the pandemic trends without face-covering using the statistical method and by projecting the trend, we calculated that over 66,000 infections were prevented by using a face mask in little over a month in New York City. We conclude that wearing a face mask in public corresponds to the most effective means to prevent inter-human transmission.

"This inexpensive practice, in conjunction with social distancing and other procedures, is the most likely opportunity to stop the COVID-19 pandemic. Our work also highlights that sound science is essential in decision-making for the current and future public health pandemics."

One of the paper's co-authors, Mario Molina, is a professor at the University of California-San Diego and a co-recipient of the 1995 Nobel Prize in Chemistry for his role in understanding the threat to the Earth's ozone layer of human-made halocarbon gases.

"Our study establishes very clearly that using a face mask is not only useful to prevent infected coughing droplets from reaching uninfected persons, but is also crucial for these uninfected persons to avoid breathing the minute atmospheric particles (aerosols) that infected people emit when talking and that can remain in the atmosphere tens of minutes and can travel tens of feet," Molina said.

Zhang said that many people in China have worn face masks for years, mainly because of the bad air quality of the country.

"So people there are sort of used to this," he said. "Mandated face-covering helped China in containing the COVID-19 outbreak."

Zhang said the results should send a clear message to people worldwide -- wearing a face mask is essential in fighting the virus.

"Our work suggests that the failure in containing the propagation of COVID-19 pandemic worldwide is largely attributed to the unrecognized importance of airborne virus transmission," he said. "Social-distancing and washing our hands must continue, but that's not sufficient enough protection. Wearing a face mask as well as practicing good hand hygiene and social distancing will greatly reduce the chances of anyone contracting the COVID-19 virus."