Lifeboat Foundation

The high energy consumption of artificial neural networks’ learning activities is one of the biggest hurdles for the broad use of Artificial Intelligence (AI), especially in mobile applications. One approach to solving this problem can be gleaned from knowledge about the human brain.

Although it has the computing power of a supercomputer, it only needs 20 watts, which is only a millionth of the energy of a supercomputer.

One of the reasons for this is the efficient transfer of information between neurons in the brain. Neurons send short electrical impulses (spikes) to other neurons—but, to save energy, only as often as absolutely necessary.

New machine learning training approach could help under-resourced academic labs catch up with big tech.

A new device that relies on flowing clouds of ultracold atoms promises potential tests of the intersection between the weirdness of the quantum world and the familiarity of the macroscopic world we experience every day. The atomtronic Superconducting QUantum Interference Device (SQUID) is also potentially useful for ultrasensitive rotation measurements and as a component in quantum computers.

“In a conventional SQUID, the quantum interference in electron currents can be used to make one of the most sensitive magnetic field detectors,” said Changhyun Ryu, a physicist with the Material Physics and Applications Quantum group at Los Alamos National Laboratory. “We use neutral atoms rather than charged electrons. Instead of responding to magnetic fields, the atomtronic version of a SQUID is sensitive to mechanical rotation.”

Although small, at only about 10 millionths of a meter across, the atomtronic SQUID is thousands of times larger than the molecules and atoms that are typically governed by the laws of quantum mechanics. The relatively large scale of the device lets it test theories of macroscopic realism, which could help explain how the world we are familiar with is compatible with the quantum weirdness that rules the universe on very small scales. On a more pragmatic level, atomtronic SQUIDs could offer highly sensitive rotation sensors or perform calculations as part of quantum computers.

A team led by Prof. Du Jiangfeng, Prof. Shi Fazhan, and Prof. Wang Ya from University of Science and Technology of China, of the Chinese Academy of Sciences, proposed a robust electrometric method utilizing a continuous dynamic decoupling technique, where the continuous driving fields provide a magnetic-field-resistant dressed frame. The study was published in Physical Review Letters on June 19.

Characterization of electrical properties and comprehension of the dynamics in nanoscale become significant in the development of modern electronic devices, such as semi-conductor transistors and quantum chips, especially when the feature size has shrunk to several nanometers.

The nitrogen-vacancy (NV) center in diamond—an atomic-scale spin sensor—has shown to be an attractive electrometer. Electrometry using the NV center would improve various sensing and imaging applications. However, its natural susceptibility to the magnetic field hinders effective detection of the electric field.

Could be used for quasar propulsion: 3.

It seems the universe has an odd sense of humor. While a crown-encrusted virus has run roughshod over the world, another entirely different corona about 100 million light years from Earth has mysteriously disappeared.

For the first time, astronomers at MIT and elsewhere have watched as a supermassive black hole’s own corona, the ultrabright, billion-degree ring of high-energy particles that encircles a black hole’s event horizon, was abruptly destroyed.

Continue reading “In a first, astronomers watch a black hole’s corona disappear, then reappear” »

At long last, the JEDEC Solid State Technology Association has officially finalized the specification for DDR5 SDRAM, which will serve as a blueprint for memory makers and future CPU platform designs as the industry shifts away from DDR4. However, it won’t happen overnight. More on that in a moment.

JEDEC is the industry group that develops and sets open memory standards, and it has more than 300 members, including AMD, Intel, Micron, Samsung, and a host of other major players in the semiconductor industry. JEDEC’s job is important because without the recognized standards that emerge, we would potentially have to navigate through a proprietary minefield when building a PC.

🤔 The WHO Chief is urging African nations to take part in clinical trials for a vaccine, yet is sending a team to China 7 months after the fact, and after China destroyed samples. The same WHO also is complaining about countries handling of the outbreak.

The UN body announced that over 1.4 million infections of COVID-19 are accounted for by health care sector workers, at least 10% of all cases. EU leaders have met to discuss a recovery package. Follow DW for the latest.

AUSTIN, Texas — Researchers in the Cockrell School of Engineering at The University of Texas at Austin have built a new type of battery that combines the many benefits of existing options while eliminating their key shortcomings and saving energy.

Most batteries are composed of either solid-state electrodes, such as lithium-ion batteries for portable electronics, or liquid-state electrodes, including flow batteries for smart grids. The UT researchers have created what they call a “room-temperature all-liquid-metal battery,” which includes the best of both worlds of liquid- and solid-state batteries.

Solid-state batteries feature significant capacity for energy storage, but they typically encounter numerous problems that cause them to degrade over time and become less efficient. Liquid-state batteries can deliver energy more efficiently, without the long-term decay of sold-state devices, but they either fall short on high energy demands or require significant resources to constantly heat the electrodes and keep them molten.

Black holes don’t glow — in fact, they’re famous for doing the opposite. But if they’re actively devouring material from the space around them, that material can blaze like a billion X-ray Suns.

And for the first time, astronomers have now seen that blaze mysteriously snuffed out, before gradually returning to brightness.

The supermassive black hole is a beast clocking in at 19 million solar masses, powering a galactic nucleus 275 million light-years away, in a galaxy called 1ES 1927+654.