Friday, February 07, 2025

Toward sustainable computing: Energy-efficient memory innovation

SOT-MRAM memory technology could replace cache memory in computer architecture in the future

Johannes Gutenberg Universitaet Mainz

How much energy is consumed each time we upload an image to social media, which relies on data centers and cloud storage? Data centers currently account for about one percent of global energy consumption, amounting to 200 terawatt-hours of electricity annually. This immense energy demand has driven researchers to explore innovative ways to reduce energy usage.

New approach is equally suitable for smartphones and supercomputers

A team of scientists at Johannes Gutenberg University Mainz (JGU) in Germany has now achieved a groundbreaking advancement in memory technology in close collaboration with Antaios, a magnetic random access memory company in France. Their innovation, based on Spin-Orbit Torque (SOT) Magnetic Random-Access Memory (MRAM), offers a highly efficient and powerful solution for data processing and storage—a transformative step forward for technologies ranging from smartphones to supercomputers.

"This prototype is one of a kind and could revolutionize data storage and processing. It aligns with global goals to reduce energy consumption and paves the way for faster, more efficient memory solutions," said Dr. Rahul Gupta, a former postdoctoral researcher at the JGU Institute of Physics, where he supervised the research, and the lead author of the study recently published in Nature Communications.

SOT-MRAM stands out for its superior power efficiency, nonvolatility, and performance compared to static RAM, making it a strong candidate to replace cache memory in computer architecture, for example. This cutting-edge technology uses electrical currents to switch magnetic states, enabling reliable data storage. However, one key challenge has been to reduce the high input current required during the writing process while ensuring industrial compatibility. This includes maintaining sufficient thermal stability to store the data for over ten years and minimizing the energy required to perform the storage task.

By exploiting previously neglected orbital currents, researchers at JGU and Antaios have developed a unique magnetic material incorporating elements such as Ruthenium as a SOT channel—a fundamental building block of SOT MRAM—to significantly enhance performance. Their innovation includes:

an over 50 percent reduction in overall energy consumption compared to existing memory technologies on an industrial scale;
a 30 percent boost in efficiency, enabling faster and more reliable data storage;
a 20 percent reduction in the input current required for magnetic switching to store the data;
the achievement of a thermal stability factor that ensures data storage longevity of more than 10 years.

The secret behind efficient memory

The breakthrough leverages a phenomenon known as the Orbital Hall Effect (OHE), enabling greater energy efficiency without relying on rare or expensive materials. Traditionally, SOT-MRAM relied on the spin property of electrons, where charge current is converted into spin current via the Spin Hall Effect. This process requires elements with high spin-orbit coupling, typically rare and expensive, often environmentally unfriendly, high atomic number materials such as platinum and tungsten. "In contrast, our approach harnesses a novel fundamental phenomenon by utilizing orbital currents derived from charge currents through the Orbital Hall Effect, eliminating the dependency on costly and rare materials," explained Dr. Rahul Gupta.

Dr. Gupta further explained that by combining this innovative approach with state-of-the-art engineering, the team has developed a scalable and practical solution ready for integration into everyday technology. This research exemplifies how scientific advancements can address some of the most pressing challenges of our time. With global energy consumption steadily increasing, breakthroughs like this highlight the crucial role of technology in creating a more sustainable future.

Successful industrial collaboration

JGU project coordinator Professor Mathias Kläui emphasized his excitement about the successful collaboration with the team of Dr. Marc Drouard at Antaios in France: "I am delighted that this collaborative effort has resulted in this exciting device concept, which is not only fascinating from a basic science point of view but might have implications in industry for GreenIT." He continued: "Reducing power consumption by discovering innovative physical mechanisms that allow for the development of more efficient technologies is one of the aims of our research."

The study was recently published in Nature Communications and has been supported by the industrial partner Antaios, the EU Research and Innovation program Horizon 2020 and Horizon Europe, the European Research Council, the German Research Foundation (DFG), and the Norwegian Research Council.

Journal

Nature Communications

DOI

10.1038/s41467-024-55437-x

Article Title

Harnessing orbital Hall effect in spin-orbit torque MRAM

Earth saw record-high greening in 2020. What’s at the root?

Duke University

As pandemic lockdowns forced humans into isolation, Earth’s vegetation was thriving. The year 2020 was the greenest in modern satellite records from 2001 to 2020, according to a recent study published in Remote Sensing of Environment. Consistent growth in northern and temperate regions, combined with a brief period of tropical growth, primarily led to this remarkably verdant period.

Terrestrial, or land-based, vegetation is vital to life on Earth as we know it. Plants regulate carbon and water cycles, feed organisms and help offset fossil fuel emissions, among many other benefits. Monitoring greening trends and variability can inform ecosystem management practices and strategies for addressing climate warming.

“Terrestrial vegetation, like trees and shrubs, take up carbon dioxide through photosynthesis, so they can offer a natural climate-warming solution,” said Yulong Zhang, a research scientist in the Division of Earth and Climate Science within the Duke University Nicholas School of the Environment, who led the multi-institutional study.

The researchers analyzed satellite data collected from 2001 to 2020 for signs of greening based on several measures, such as vegetation structure, density and plant health. An overall trend toward more abundant vegetation emerged, with considerable variability from year to year. However, 2020 stood out as markedly greener than the others.

The team wondered: Was the pandemic responsible for the leafy bonanza? For example, lockdowns and travel restrictions temporarily reduced daily activity and air pollution, which could have enabled plants to flourish in brighter sunlight in the absence of human disturbance, the authors suggested.

To tease out the drivers of Earth’s green-up, they used machine learning and ecological simulations to explore how different environmental and human factors might have affected vegetation growth. Contrary to expectations, pandemic-related lockdowns had a limited effect on global greening.

“Although short-term air quality improvements and reduced disturbances slightly boosted greenness in lockdown regions, these effects faded quickly and were further counteracted by natural climate variability as the year progressed,” Zhang said.

The team found that 2020’s record-high greening was instead primarily associated with temperate and colder regions, fueled by rising CO2 levels, climate warming and reforestation efforts. The effects of climate warming were most prominent in cooler areas, such as boreal forests, where rising temperatures extended growing seasons.

Additionally, “China and India were massive green hotspots, thanks to tree planting, land restoration and agricultural management efforts,” noted co-author Tong Qiu of the Nicholas School. “High rainfall in 2019-2020 also boosted vegetation growth in the tropics, likely linked with climate oscillations, or patterns, such as La Niña and the Indian Ocean Dipole.”

The study highlights Earth’s remarkable ability to adapt to environmental changes, the authors concluded. However, they issued a note of caution.

“Climate extremes, water scarcity, wildfire risk and increased human pressures could potentially slow down or even reverse these gains under a warming climate,” said co-author Wenhong Li, also of the Nicholas School. “If that happens, we may see an acceleration of the ongoing climate warming.”

To that end, the team is calling for expanded monitoring and the development of advanced models to predict vegetation trends and variability as climate change accelerates.

Citation: “Earth’s record-high greenness and its attributions in 2020,” Yulong Zhang, Jiafu Mao, Ge Sun, Qinfeng Guo, Jeffrey Atkins, Wenhong Li, Mingzhou Jin, Conghe Song, Jingfeng Xiao, Taehee Hwang, Tong Qiu, Lin Meng, Daniel M. Ricciuto, Xiaoying Shi, Xing Li, Peter Thornton, Forrest Hoffman. Remote Sensing of Environment, Jan. 1, 2025, DOI: 10.1016/j.rse.2024.114494

Online: https://www.sciencedirect.com/science/article/pii/S0034425724005200

# # #

Journal

Remote Sensing of Environment

DOI

10.1016/j.rse.2024.114494

Method of Research

Computational simulation/modeling

Subject of Research

Not applicable

Article Title

Earth’s record-high greenness and its attributions in 2020

Ice streams move due to tiny ice quakes

Dynamics of Greenland’s ice decrypted

News Release 6-Feb-2025
ETH Zurich
Facebook
X LinkedIn WeChat Bluesky Message WhatsApp Email

image:
A researcher lowers a fibre-optic cable 1,500 metres into the borehole in or-der to record signals from inside the ice stream continuously for 14 hours.
view more
Credit: ETH Zurich

In brief:
•             In Greenland, an international team of researchers led by ETH Zurich has discovered that countless tiny ice quakes take place deep inside ice streams.
•             These quakes are responsible for the fact that ice streams also move with a continuous stick-slip motion and not only like viscous honey as previously considered.
•             The researchers recorded seismic data from inside the ice stream using a fibre-optic cable in a 2,700-metre deep borehole.

The great ice streams of the Antarctic and Greenland are like frozen rivers, carrying ice from the massive inland ice sheets to the sea – and a change in their dynamics will contribute significantly to sea-level rise. In order to estimate just how much sea levels will rise, climate researchers rely on computer simulations of the ice streams. Until now, they have based these simulations on an assumption that the ice streams flow slowly but steadily into the sea like thick honey.
However, satellite measurements of the flow speed of ice streams show that such simulations are inaccurate and have shortcomings to correctly reflect reality. This leads to considerable uncertainties in estimates of how much mass the ice streams are losing and how quickly and how high sea levels will rise.
Ice streams both judder and flow
Now, a team of researchers led by ETH professor Andreas Fichtner has made an unexpected discovery: deep within the ice streams, there are countless weak quakes taking place that trigger one another and propagate over distances of hundreds of metres. This discovery helps to explain the discrepancy between current simulations of ice streams and satellite measurements, and the new findings should also impact the way ice streams are simulated in the future.
“The assumption that ice streams only flow like viscous honey is no longer tenable. They also move with a constant stick-slip motion,” says Fichtner. The ETH professor is confident that this finding will be integrated into simulations of ice streams, making estimates of changes in sea level more accurate.
Riddles relating to ice cores resolved
Moreover, the ice quakes explain the origin of numerous fault planes between ice crystals in ice cores obtained from great depths. These fault planes are the result of tectonic shifts and have been known to scientists for decades, although no explanation had been found for them until now.
“The fact that we’ve now discovered these ice quakes is a key step towards gaining a better understanding of the deformation of ice streams on small scales,” explains Olaf Eisen, Professor at the Alfred Wegener Institute and one of the study’s co-authors.
The study by this international research team led by ETH Zurich has just been published in the journal Science and also involved researchers from the Alfred Wegener Institute, Helmholtz Centre for Polar and Marine Research (AWI), the University of Strasbourg, the Niels Bohr Institute (NBI), the Swiss Federal Institute WSL and other universities.
Fire and ice are related
The fact that these ice quakes cannot be observed at the surface and have therefore remained undiscovered until now is due to a layer of volcanic particles located 900 metres below the surface of the ice. This layer stops the quakes from propagating to the surface. Analysis of the ice core showed that these volcanic particles originate from a massive eruption of Mount Mazama in what is now Oregon (USA) some 7,700 years ago. “We were astonished by this previously unknown relationship between the dynamics of an ice stream and volcanic eruptions,” Fichtner recalls.
The ETH professor also noticed that the ice quakes start from impurities in the ice. These impurities are also leftovers from volcanoes: tiny traces of sulphates that entered the atmosphere in volcanic eruptions and flew halfway around the world before being deposited on the Greenland ice sheet in snowfall. These sulphates reduce the stability of the ice and favour the formation of microfissures.
A 2,700-metre borehole in the ice
The researchers discovered the ice quakes using a fibre-optic cable that was inserted into a 2,700-metre-deep borehole and recorded seismic data from inside a massive ice stream for the first time. This borehole was drilled into the ice by researchers from the East Greenland Ice-core Project (EastGRIP), led by the Niels Bohr Institute and strongly supported by the Alfred Wegener Institute, resulting in the extraction of a 2,700-metre-long ice core. Once drilling work was complete, the researchers took the opportunity to lower a fibre-optic cable 1,500 metres into the borehole and record signals from inside the ice stream continuously for 14 hours.
The research station and borehole are located on the North East Greenland Ice Stream (NEGIS), around 400 kilometres from the coast. The NEGIS is the biggest ice stream of the Greenland ice sheet, whose retreat is a large contributor to current rising sea levels. In the area of the research station, the ice is moving towards the sea at a speed of around 50 metres per year.
As ice quakes occur frequently over a wide area in the researchers’ measurements, ETH researcher Fichtner believes it is also plausible that they occur in ice streams everywhere, all the time. To verify this, however, it will be necessary to take seismic measurements of this kind in other boreholes – and there are already plans to do just that.
###
References
Fichtner A, Hofstede C, Kennett B L N, Svensson A, Westhoff J, Walter F, Ampuero J-P, Cook E, Zigone D, Jansen D, Eisen O, Hidden cascades of seismic ice stream deformation. Science. 6.2.2025, DOI: 10.1126/science.adp8094

Journal

Science

DOI

10.1126/science.adp8094

Article Title

Hidden cascades of seismic ice stream deformation

Article Publication Date

6-Feb-2025

Special Issue: The cryosphere

Summary author: Walter Beckwith

American Association for the Advancement of Science (AAAS)

In this Special Issue of Science, 3 Reviews and a Policy Forum highlight research on Earth’s frozen places – from the Arctic to the Antarctic – and how it’s changing due to climate change and the geopolitical challenges this important work faces. In the first Review, Julienne Stroeve and colleagues provide a preview of what the Arctic region may look like in a warmer world. Without stronger climate action, global temperatures are set to rise +2.7°C above preindustrial levels, causing irreversible Arctic transformation. Under this future, Stroeve et al. show that nearly all days in the region would exceed past temperature extremes, summers would see an ice-free Arctic Ocean, Greenland’s melt zones would quadruple, and permafrost would shrink by half. While these changes would severely disrupt ecosystems, and damage Arctic communities and infrastructure, the authors note that stronger climate action could significantly mitigate these consequences, preserving the Arctic’s stability and resilience. In a second Review, Helen Fricker and colleagues focus on Antarctic ice, which plays a crucial role in regulating Earth’s climate, global sea levels, ocean circulation, and planetary reflectivity. Fricker et al. highlight key knowledge gaps in these processes, which create uncertainties in the stability of the Antarctic ice sheet and the potential impacts of its continued loss. According to Fricker et al., progress in these areas depends on high-resolution satellite monitoring, targeted field studies, improved modeling, and strengthened interdisciplinary collaboration to reduce uncertainties and refine future projections. A third Review focuses on Antarctic biodiversity. According to Luis Pertierra and colleagues, Antarctica hosts myriad unique life forms, yet much of the region’s biodiversity and ecological functioning remains poorly understood. Outside of a few well-known vertebrate species, much about the region’s invertebrate, plant, and microbial life remains unknown. Here, Pertierra et al. highlight the shortfalls in Antarctic biodiversity research, explain how these gaps hinder ecological insight and conservation in the region, and offer a framework for addressing them. Lastly, in a Policy Forum, Jennifer Spence and colleagues discuss the geopolitical challenges facing Arctic research. The Arctic is at the forefront of critical global issues, including the uneven impacts of climate change, the challenge of balancing economic development with conservation, and the complexity of governance across multiple boundaries. These issues are complicated further by rising populism, geopolitical tensions, and distrust in governmental institutions. Spence et al. highlight three key issues, including understanding the regional and global effects of climate change, ensuring research benefits Arctic communities, and maintaining international cooperation in the face of growing geopolitical pressures. However, despite these challenges, the Arctic region has pioneered innovative governance models that prioritize Indigenous engagement, as well as fostering international cooperation through institutions like the Arctic Council to promote sustainable development and environmental protection.

Journal

Science

DOI

10.1126/science.adw2518

Article Title

Extreme conditions

Article Publication Date

7-Feb-2025

Scientists discover brain mechanism that helps overcome fear

Research reveals how the brain learns to suppress instinctive fear responses, pointing to new potential targets for PTSD and anxiety treatments

Sainsbury Wellcome Centre

Brain slice showing projections to the vLGN — image:
Coronal brain slice showing projections from different visual areas in the cerebral cortex to the ventrolateral geniculate nucleus (vLGN). These pathways are part of the circuit identified as mediating the suppression of instinctive fear responses.
view more
Credit: Sainsbury Wellcome Centre

Researchers at the Sainsbury Wellcome Centre (SWC) at UCL have unveiled the precise brain mechanisms that enable animals to overcome instinctive fears. Published today in Science, the study in mice could have implications for developing therapeutics for fear-related disorders such as phobias, anxiety and post-traumatic stress disorder (PTSD).

The research team, led by Dr Sara Mederos and Professor Sonja Hofer, mapped out how the brain learns to suppress responses to perceived threats that prove harmless over time.

"Humans are born with instinctive fear reactions, such as responses to loud noises or fast-approaching objects," explains Dr Mederos, Research Fellow in the Hofer Lab at SWC. "However, we can override these instinctive responses through experience – like children learning to enjoy fireworks rather than fear their loud bangs. We wanted to understand the brain mechanisms that underlie such forms of learning”.

Using an innovative experimental approach, the team studied mice presented with an overhead expanding shadow that mimicked an approaching aerial predator. Initially, the mice sought shelter when encountering this visual threat. However, with repeated exposure and no actual danger, the mice learned to remain calm instead of escaping, providing researchers with a model to study the suppression of fear responses.

Based on previous work in the Hofer Lab, the team knew that an area of the brain called the ventrolateral geniculate nucleus (vLGN) could suppress fear reactions when active and was able to track knowledge of previous experience of threat. The vLGN also receives strong input from visual areas in the cerebral cortex, and so the researchers explored whether this neural pathway had a role in learning not to fear a visual threat.

The study revealed two key components in this learning process: (1) specific regions of the visual cortex proved essential for the learning process, and (2) a brain structure called the ventrolateral geniculate nucleus (vLGN) stores these learning-induced memories.

“We found that animals failed to learn to suppress their fear responses when specific cortical visual areas where inactivated. However, once the animals had already learned to stop escaping, the cerebral cortex was no longer necessary,” explained Dr Mederos.

"Our results challenge traditional views about learning and memory," notes Professor Hofer, senior author of the study. "While the cerebral cortex has long been considered the brain's primary centre for learning, memory and behavioural flexibility, we found the subcortical vLGN and not the visual cortex actually stores these crucial memories. This neural pathway can provide a link between cognitive neocortical processes and ‘hard-wired’ brainstem-mediated behaviours, enabling animals to adapt instinctive behaviours.”

The researchers also uncovered the cellular and molecular mechanisms behind this process. Learning occurs through increased neural activity in specific vLGN neurons, triggered by the release of endocannabinoids – brain-internal messenger molecules known to regulate mood and memory. This release decreases inhibitory input to vLGN neurons, resulting in heightened activity in this brain area when the visual threat stimulus is encountered, which suppresses fear responses.

The implications of this discovery extend beyond the laboratory. “Our findings could also help advance our understanding of what is going wrong in the brain when fear response regulation is impaired in conditions such as phobias, anxiety and PTSD. While instinctive fear reactions to predators may be less relevant for modern humans, the brain pathway we discovered exists in humans too," explains Professor Hofer. "This could open new avenues for treating fear disorders by targeting vLGN circuits or localised endocannabinoid systems."

The research team is now planning to collaborate with clinical researchers to study these brain circuits in humans, with the hope of someday developing new, targeted treatments for maladaptive fear responses and anxiety disorders.

This research was funded by the Sainsbury Wellcome Centre core grant from the Gatsby Charity Foundation and Wellcome (090843/F/09/Z); a Wellcome Investigator Award (219561/Z/19/Z); an EMBO postdoctoral fellowship (EMBO ALTF 327-2021) and a Wellcome Early Career Award (225708/Z/22/Z).

Source:

Read the full paper in Science: ‘Overwriting an instinct: visual cortex instructs learning to suppress fear responses’

Media contact:

For more information or to speak to the researchers involved, please contact:

April Cashin-Garbutt, Head of Research Communications and Engagement, Sainsbury Wellcome Centre
E: a.cashin-garbutt@ucl.ac.uk T: +44 (0)20 3108 8028

About the Sainsbury Wellcome Centre

The Sainsbury Wellcome Centre (SWC) brings together world-leading neuroscientists to generate theories about how neural circuits in the brain give rise to the fundamental processes underpinning behaviour, including perception, memory, expectation, decisions, cognition, volition and action. Funded by the Gatsby Charitable Foundation and Wellcome, SWC is located within UCL and is closely associated with the Faculties of Life Sciences and Brain Sciences. For further information, please visit: www.sainsburywellcome.org

Journal

Science

DOI

10.1126/science.adr2247

Method of Research

Experimental study

Subject of Research

Animals

Article Title

Overwriting an instinct: Visual cortex instructs learning to suppress fear responses

Article Publication Date

7-Feb-2025

Disclaimer: AAAS a

LA REVUE GAUCHE - Left Comment

Friday, February 07, 2025

Toward sustainable computing: Energy-efficient memory innovation

Journal

DOI

Article Title

Earth saw record-high greening in 2020. What’s at the root?

Journal

DOI

Method of Research

Subject of Research

Article Title

Ice streams move due to tiny ice quakes

Journal

Science

DOI

10.1126/science.adp8094

Article Title

Hidden cascades of seismic ice stream deformation

Article Publication Date

6-Feb-2025

Special Issue: The cryosphere

Journal

DOI

Article Title

Article Publication Date

Scientists discover brain mechanism that helps overcome fear

Journal

DOI

Method of Research

Subject of Research

Article Title

Article Publication Date

About Me