Nvidia is powering the AI revolution, but investor Cathie Wood warns there’s a clear risk its customers will cut orders and serious competition will emerge.
This article is part of our coverage of the latest in AI research.
Diffusion models are best known for their impressive capabilities to generate highly detailed images. They are the main architecture used in popular text-to-image models such as DALL-E, Stable Diffusion, and Midjourney.
However, diffusion models can be used for more than just generating images. A new paper by researchers at Meta, Princeton University, and University of Texas, Austin, shows that diffusion models can help create better reinforcement learning systems.
New phased-array transmitter design overcomes common problems of CMOS technology in the 300 GHz band, as reported by scientists from Tokyo Tech. Thanks to its remarkable area efficiency, low power consumption, and high data rate, the proposed transmitter could pave the way to many technological applications in the 300 GHz band, including body and cell monitoring, radar, 6G wireless communications, and terahertz sensors.
Today, most frequencies above the 250 GHz mark remain unallocated.
Accordingly, many researchers are developing 300 GHz transmitters/receivers to capitalize on the low atmospheric absorption at these frequencies, as well as the potential for extremely high data rates that comes with it.
Researchers at the University of Colorado Anschutz Medical Campus have found that inhibiting a key protein can stop the destruction of synapses and dendritic spines commonly seen in Alzheimer’s disease.
The study, whose first author is Tyler Martinez, a student in the Pharmacology and Molecular Medicine PhD program at the University of Colorado School of Medicine, was published recently in the journal eNeuro.
The researchers, using rodent neurons, found that targeting a protein called Mdm2 with an experimental cancer drug known as nutlin, stopped neurotoxic amyloid-b peptides that accumulate in Alzheimer’s disease (AD) from overly pruning synapses.
Dark matter comprises around 85% of all the matter in the universe. Although ordinary matter absorbs, reflects and emits light, dark matter cannot be seen directly, which makes its detection difficult. Its existence is inferred from its gravitational effects on visible matter, the material that forms stars, planets, and other objects in the cosmos.
Galaxies are made up of these two types of material The dark matter is distributed in halos, which are huge structures surrounding galaxies, while the ordinary matter is mainly present in the central regions where most of the stars are found.
Traditionally observational studies of galactic evolution have centered on the role of ordinary matter, even though it is quite a small fraction of the mass of a galaxy. For decades there have been theoretical predictions about the effect that dark matter should have on the evolution of galaxies. However, in spite of numerous efforts, there is no clear consensus about this.
European astronomers have conducted very long baseline interferometric (VLBI) observations of a radio jet structure in a powerful quasar known as PKS 2215+020. The collected VLBI data provide important insights into the properties of this jet, suggesting that PKS 2215+020 is a blazar. The findings were presented February 17 in the Universe journal.
Quasars, or quasi-stellar objects (QSOs) are active galactic nuclei (AGN) of very high luminosity, emitting electromagnetic radiation observable in radio, infrared, visible, ultraviolet and X-ray wavelengths.
They are among the brightest and most distant objects in the known universe, and serve as fundamental tools for numerous studies in astrophysics as well as cosmology. For instance, quasars have been used to investigate the large-scale structure of the universe and the era of reionization. They also improve our understanding of the dynamics of supermassive black holes and the intergalactic medium.
A long-term analysis conducted by leading microbiologists at the Icahn School of Medicine at Mount Sinai reveals that antibody responses induced by COVID-19 vaccines are long-lasting. The study results, published online in the journal Immunity challenge the idea that mRNA-based vaccine immunity wanes quickly.
The emergence of SARS-CoV-2, the virus that causes COVID-19, in late 2019 sparked the global pandemic that is now in its fifth year. Vaccines that were developed at record speed have saved millions of lives. However, the emergence of SARS-CoV-2 variants and waning immunity have decreased the effectiveness of the vaccines against symptomatic disease. The common perception now is that mRNA-based vaccine-induced immunity wanes quickly.
However, this assumption is largely based on data from short-term studies that include a very limited number of data points following peak responses.
In an article published in the Journal of Materials Chemistry C, Brazilian researchers describe a strategy to enhance the efficiency and stability of solar cells made of perovskite, a semiconductor material produced in the laboratory. The results of the project could be highly positive for the future of the solar power sector.
Developed by researchers at São Paulo State University (UNESP) in Bauru, Brazil, the method involves the use of a class of materials known as MXenes, a family of two-dimensional materials with a graphene-like structure combining transition metals, carbon and/or nitrogen, and surface functional groups such as fluoride, oxygen or hydroxyl. Their properties include high electrical conductivity, good thermal stability, and high transmittance (relating to the amount of light that passes through a substance without being reflected or absorbed).
In the study, the MXene Ti3C2Tx was added to polymethyl methacrylate (PMMA) to form a passivation coating, which was spin-coated on top of the perovskite layer of inverted solar cells. Passivation coatings are designed to mitigate possible defects in polycrystalline solids (perovskite in this case) due to interaction with the environment or to their internal structure.
Science: In future maybe wounds be cured and closed in seconds by 3D printing regeneration.
Fat tissue holds the key to 3D printing layered living skin and potentially hair follicles, according to researchers who recently harnessed fat cells and supporting structures from clinically procured human tissue to precisely correct injuries in rats. The advancement could have implications for reconstructive facial surgery and even hair growth treatments for humans.
The team’s findings were published March 1 in Bioactive Materials. The U.S. Patent and Trademark Office granted the team a patent in February for the bioprinting technology it developed and used in this study.
Continue reading “3D-printed skin closes wounds and contains hair follicle precursors” »
According to a Mayo Clinic study published in Nature Neuroscience, the cells that act as the central nervous system’s first line of defense against harm also play a role in helping the brain awaken from anesthesia. This discovery could help pave the way for innovative methods that address post-anesthesia complications.
When coming out of anesthesia, more than one-third of patients can experience either extreme drowsiness or hyperactivity, a side effect called delirium. Mayo researchers found that special immune cells in the brain called microglia can act to shield neurons from the aftereffects of anesthesia to awaken the brain.
“This is the first time we’ve seen microglia enhance and boost neuronal activity by physically engaging the brain circuits,” says Mayo Clinic neuroscientist Long-Jun Wu, senior author of the study.
