Now, researchers at Princeton Engineering and the Indian Institute of Technology have harnessed artificial intelligence to take a key step toward slashing the time and cost of designing new wireless chips and discovering new functionalities to meet expanding demands for better wireless speed and performance. In an article published Dec. 30 in Nature Communications, the researchers describe their methodology, in which an AI creates complicated electromagnetic structures and associated circuits in microchips based on the design parameters. What used to take weeks of highly skilled work can now be accomplished in hours.

What is more, the AI behind the new system has produced strange new designs featuring unusual patterns of circuitry. Kaushik Sengupta, the lead researcher, said the designs were unintuitive and unlikely to be developed by a human mind. But they frequently offer marked improvements over even the best standard chips.

“We are coming up with structures that are complex and looks random shaped and when connected with circuits, they create previously unachievable performance. Humans cannot really understand them, but they can work better,” said Sengupta, a professor of electrical and computer engineering and co-director of NextG, Princeton’s industry partnership program to develop next-generation communications.

These circuits can be engineered towards more energy efficient operation or to make them operable across an enormous frequency range that is not currently possible. Furthermore, the method synthesizes inherently complex structures in minutes, while conventional algorithms may take weeks. In some cases, the new methodology can create structures that are impossible to synthesize with current techniques.

Uday Khankhoje, a co-author and associate professor of electrical engineering at IIT Madras, said the new technique not only delivers efficiency but promises to unlock new approaches to design challenges that have been beyond the capability of engineers.

“This work presents a compelling vision of the future,” he said. “AI powers not just the acceleration of time-consuming electromagnetic simulations, but also enables exploration into a hitherto unexplored design space and delivers stunning high-performance devices that run counter to the usual rules of thumb and human intuition.”

Wireless chips are a combination of standard electronic circuits like those in computer chips and electromagnetic structures including antennas, resonators, signal splitters, combiners and others. These combinations of elements are put together in every circuit block, carefully handcrafted and co-designed to operate optimally. This method is then scaled to other circuits, sub-systems and systems, making the design process extremely complex and time consuming, particularly for modern, high-performance chips behind applications like wireless communication, autonomous driving, radar and gesture recognition.

“Classical designs, carefully, put these circuits and electromagnetic elements together, piece by piece, so that the signal flows in the way we want it to flow in the chip. By changing those structures, we incorporate new properties,” Sengupta said. “Before, we had a finite way of doing this, but now the options are much larger.”

It can be hard to comprehend the vastness of a wireless chip’s design space. The circuitry in an advanced chip is so small, and the geometry so detailed, that the number of possible configurations for a chip exceeds the number of atoms in the universe, Sengupta said. There is no way for a person to understand that level of complexity, so human designers don’t try. They build chips from the bottom up, adding components as needed and adjusting the design as they build.

The AI approaches the challenge from a different perspective, Sengupta said. It views the chip as a single artifact. This can lead to strange, but effective arrangements. He said humans play a critical role in the AI system, in part because that AI can make faulty arrangements as well as efficient ones. It is possible for AI to hallucinate elements that don’t work, at least for now. This requires some level of human oversight.

“There are pitfalls that still require human designers to correct,” Sengupta said. “The point is not to replace human designers with tools. The point is to enhance productivity with new tools. The human mind is best utilized to create or invent new things, and the more mundane, utilitarian work can be offloaded to these tools.”

The researchers have used AI to discover and design complex electromagnetic structures that are-co-designed with circuits to create broadband amplifiers. Sengupta said future research will involve linking multiple structures and designing entire wireless chips with the AI system.

“Now that this has shown promise, there is a larger effort to think about more complicated systems and designs,” he said. “This is just the tip of the iceberg in terms of what the future holds for the field.”

In the elderly, however, excessive filtration cannot be detected due to age-related GFR decline. To accurately assess GFR, Osaka Metropolitan University researchers have come up with a new calculation method.

The group led by Dr. Akihiro Tsuda, a lecturer at the Graduate School of Medicine, assessed 180 kidney transplant donor candidates to define a new formula for determining the threshold value for hyperfiltration based on age and GFR values.

Among other findings, the conventional method of correcting for body surface area in obese patients was determined to be inaccurate as excessive filtration cannot be detected. The researchers suggest calculating GFR without the correction but by taking into account the decline in the filtration rate due to aging.

“Since hyperfiltration is a precursor to diabetic nephropathy, we hope that using this new formula will more accurately diagnose the condition, leading to early detection and treatment,” stated Dr. Tsuda.

The findings were published in Hypertension Research.

A team of scientists based in the U.S. and Canada recently confirmed that carbon and other star-formed atoms don’t just drift idly through space until they are dragooned for new uses. For galaxies like ours, which are still actively forming new stars, these atoms take a circuitous journey. They circle their galaxy of origin on giant currents that extend into intergalactic space. These currents — known as the circumgalactic medium — resemble giant conveyer belts that push material out and draw it back into the galactic interior, where gravity and other forces can assemble these raw materials into planets, moons, asteroids, comets and even new stars.

“Think of the circumgalactic medium as a giant train station: It is constantly pushing material out and pulling it back in,” said team member Samantha Garza, a University of Washington doctoral candidate. “The heavy elements that stars make get pushed out of their host galaxy and into the circumgalactic medium through their explosive supernovae deaths, where they can eventually get pulled back in and continue the cycle of star and planet formation.”

Garza is lead author on a paper describing these findings that was published Dec. 27 in the Astrophysical Journal Letters.

“The implications for galaxy evolution, and for the nature of the reservoir of carbon available to galaxies for forming new stars, are exciting,” said co-author Jessica Werk, UW professor and chair of the Department of Astronomy. “The same carbon in our bodies most likely spent a significant amount of time outside of the galaxy!”

In 2011, a team of scientists for the first time confirmed the long-held theory that star-forming galaxies like ours are surrounded by a circumgalactic medium — and that this large, circulating cloud of material includes hot gases enriched in oxygen. Garza, Werk and their colleagues have discovered that the circumgalactic medium of star-forming galaxies also circulates lower-temperature material like carbon.

“We can now confirm that the circumgalactic medium acts like a giant reservoir for both carbon and oxygen,” said Garza. “And, at least in star-forming galaxies, we suggest that this material then falls back onto the galaxy to continue the recycling process.”

Studying the circumgalactic medium could help scientists understand how this recycling process subsides, which will happen eventually for all galaxies — even ours. One theory is that a slowing or breakdown of the circumgalactic medium’s contribution to the recycling process may explain why a galaxy’s stellar populations decline over long periods of time.

“If you can keep the cycle going — pushing material out and pulling it back in — then theoretically you have enough fuel to keep star formation going,” said Garza.

For this study, the researchers used the Cosmic Origins Spectrograph on the Hubble Space Telescope. The spectrograph measured how light from nine distant quasars — ultra-bright sources of light in the cosmos — is affected by the circumgalactic medium of 11 star-forming galaxies. The Hubble readings indicated that some of the light from the quasars was being absorbed by a specific component in the circumgalactic medium: carbon, and lots of it. In some cases, they detected carbon extending out almost 400,000 light years — or four times the diameter of our own galaxy — into intergalactic space.

Future research is needed to quantify the full extent of the other elements that make up the circumgalactic medium and to further compare how their compositions differ between galaxies that are still making large amounts of stars and galaxies that have largely ceased star formation. Those answers could illuminate not just when galaxies like ours transition into stellar deserts, but why.

Co-authors on the paper are Trystyn Berg, research fellow at the Herzberg Astronomy and Astrophysics Research Centre in British Columbia; Yakov Faerman, a UW postdoctoral researcher in astronomy; Benjamin Oppenheimer, a research fellow at the University of Colorado Boulder; Rongmon Bordoloi, assistant professor of physics at North Carolina State University; and Sara Ellison, professor of physics and astronomy at the University of Victoria. The research was funded by NASA and the National Science Foundation.

After its discovery, which revolutionized oncology and earned a 2018 Nobel Prize, researchers developed new drugs to block PD-1 and unleash the body’s immune system to fight cancer. Yet treatments leveraging PD-1 are only effective in a small fraction of cancer patients, highlighting the need for a deeper understanding of how PD-1 works. Much of our current knowledge of PD-1’s functions comes from studies in mice, grounded on the assumption that rodent and human biology operate similarly.

Researchers in UC San Diego’s School of Biological Sciences and School of Medicine have now discovered that this assumption may be flawed. In a comprehensive assessment of PD-1 that featured novel biochemical analyses, animal modeling and a new evolutionary roadmap tracing PD-1 back millions of years, the UC San Diego scientists and their colleagues at the Chinese Academy of Sciences found that PD-1 in mice is significantly weaker than the human version.

The study, led by assistant project scientist Takeya Masubuchi, revealed several previously unknown PD-1 characteristics, including a “motif” — a specific sequence of amino acids — that is vastly different in rodents and humans.

“Our work uncovers unexpected species-specific features of PD-1 with implications for developing better pre-clinical models for PD-1,” said Associate Professor Enfu Hui of the School of Biological Sciences, Department of Cell and Developmental Biology, and a senior author of the paper. “We found a motif in PD-1 that’s present in most mammals, including humans, but is surprisingly missing in rodents, making rodent PD-1 uniquely weaker.”

The results of the study are published January 3, 2025, in the journal Science Immunology.

“Although many proteins in mice and humans have similar sequences, receptors in the immune system often show greater differences,” said Masubuchi. “Our study shows that these sequence differences can lead to functional variations of immune checkpoint receptors across species.”

Furthering their analysis, the researchers tested the impact of PD-1 humanization in mice — replacing mouse PD-1 with the human version — through co-senior author Professor Jack Bui’s laboratory in the Department of Pathology. They found that PD-1 humanization disrupted the ability of immune cells (T cells) to combat tumors.

“This study shows that as science progresses we need to have a rigorous understanding of the model systems that we use to develop medicines and drugs,” said Bui. “Finding that rodents might be outliers in terms of PD-1 activity forces us to rethink how to deploy medicines to people. If we’ve been testing medicines in rodents and they’re really outliers, we might need better model systems.”

To trace the PD-1 human-rodent differences over time, the researchers collaborated with co-senior author Professor Zhengting Zou and his Chinese Academy of Sciences colleagues. They discovered evidence of a major dip in ancestral rodent PD-1 activity around 66 million years ago after the Cretaceous-Paleogene (K-Pg) mass extinction event, which wiped out the (non-avian) dinosaurs. The analysis showed that the rodent PD-1 is uniquely weak among all vertebrates. The weakening may be attributed to special ecological adaptations to escape the effects of rodent-specific pathogens.

“The rodent ancestors survived the extinction event but their immune receptor activities or landscape might have been altered as a consequence of adaptation to new environmental challenges,” said Hui.

Future studies will assess the impact of PD-1 on the anti-tumor activity of T cells in a humanized context across various tumor types.

“The extraordinary switching and memory capabilities of these nanocrystals may one day become integral to optical computing — a way to rapidly process and store information using light particles, which travel faster than anything in the universe,” said Artiom Skripka, assistant professor in the OSU College of Science. “Our findings have the potential to advance artificial intelligence and information technologies generally.”

Published in Nature Photonics, the study by Skripka and collaborators at Lawrence Berkeley National Laboratory, Columbia University and the Autonomous University of Madrid involves a type of material known as avalanching nanoparticles.

Nanomaterials are tiny bits of matter measuring between one-billionth and one-hundred-billionths of a meter, and avalanching nanoparticles feature extreme non-linearity in their light-emission properties — they emit light whose intensity can increase massively with a small increase in the intensity of the laser that’s exciting them.

The researchers studied nanocrystals composed of potassium, chlorine and lead and doped with neodymium. By themselves, the potassium lead chloride nanocrystals do not interact with light; however, as hosts, they enable their neodymium guest ions to handle light signals more efficiently, making them useful for optoelectronics, laser technology and other optical applications.

“Normally, luminescent materials give off light when they are excited by a laser and remain dark when they are not,” Skripka said. “In contrast, we were surprised to find that our nanocrystals live parallel lives. Under certain conditions, they show a peculiar behavior: They can be either bright or dark under exactly the same laser excitation wavelength and power.”

That behavior is referred to as intrinsic optical bistability.

“If the crystals are dark to start with, we need a higher laser power to switch them on and observe emission, but once they emit, they remain emitting and we can observe their emission at lower laser powers than we needed to switch them on initially,” Skripka said. “It’s like riding a bike — to get it going, you have to push the pedals hard, but once it is in motion, you need less effort to keep it going. And their luminescence can be turned on and off really abruptly, as if by pushing a button.”

The low-power switching capabilities of the nanocrystals align with the global effort to reduce the amount of energy consumed by the growing presence of artificial intelligence, data centers and electronic devices. And not only do AI applications require substantial computational power, they are often constrained by limitations associated with existing hardware, a situation this new research could also address.

“Integrating photonic materials with intrinsic optical bistability could mean faster and more efficient data processors, enhancing machine learning algorithms and data analysis,” Skripka said. “It could also mean more-efficient light-based devices of the type used in fields like telecommunications, medical imaging, environmental sensing, and interconnects for optical and quantum computers.”

Additionally, he said, the study complements existing efforts to develop powerful, general-purpose optical computers, which are based on the behavior of light and matter at the nanoscale, and underscores the importance of fundamental research in driving innovation and economic growth.

“Our findings are an exciting development, but more research is necessary to address challenges such as scalability and integration with existing technologies before our discovery finds a home in practical applications,” Skripka said.

The U.S. Department of Energy, the National Science Foundation and the Defense Advanced Research Projects Agency supported the research, which was led by Bruce Cohen and Emory Chan of Lawrence Berkeley, P. James Schuck of Columbia University and Daniel Jaque of the Autonomous University of Madrid.