A tiny little winged drone can soar with close to zero throttle, thanks to an impressive algorithm that responds to changing winds.
Ticking clocks and flashing fireflies that start out of sync will fall into sync, a tendency that has been observed for centuries. A discovery two decades ago therefore came as a surprise: the dynamics of identical coupled oscillators can also be asynchronous. The ability to fall in and out of sync, a behavior dubbed a chimera state, is generic to identical coupled oscillators and requires only that the coupling is nonlocal. Now Yasuhiro Yamada and Kensuke Inaba of NTT Basic Research Laboratories in Japan show that this behavior can be analyzed using a lattice model (the XY model) developed to understand antiferromagnetism [1]. Besides a pleasing correspondence, Yamada and Inaba say that their finding offers a path to study the partial synchronization of neurons that underlie brain function and dysfunction.
The chimera states of a system are typically analyzed by looking at how the relative phases of the coupled oscillators fall in and out of sync. But that approach struggles to describe the system when the system contains distantly separated pockets of synchrony or when there are nontrivial configurations of the oscillators, such as twisted or spiral waves. It also requires knowledge of the network’s structure and the oscillators’ equations of motion.
In seeking an alternative approach, Yamada and Inaba turned to a two-dimensional lattice model used to tackle phase transitions in 2D condensed-matter systems. A crucial ingredient in that model is a topological defect called a vortex. Yamada and Inaba found that they could embody the asynchronous dynamics of pairs of oscillators by formulating the problem in terms of an analogous quantity that they call pseudovorticity, whose absence indicates synchrony and whose presence indicates asynchrony. Their calculations show that their pseudo-vorticity-containing lattice model can successfully recover the chimera state behavior of a simulated neural network made up of 200 model oscillators of a type commonly used to study brain activity.
Wave-based analog computing has recently emerged as a promising computing paradigm due to its potential for high computational efficiency and minimal crosstalk. Although low-frequency acoustic analog computing systems exist, their bulky size makes it difficult to integrate them into chips that are compatible with complementary metal-oxide semiconductors (CMOS). This research paper addresses this issue by introducing a compact analog computing system (ACS) that leverages the interactions between ultrasonic waves and metasurfaces to solve ordinary and partial differential equations. The results of our wave propagation simulations, conducted using MATLAB, demonstrate the high accuracy of the ACS in solving such differential equations. Our proposed device has the potential to enhance the prospects of wave-based analog computing systems as the supercomputers of tomorrow.
A new study led by University of Maryland physicists sheds light on the cellular processes that regulate genes. Published in the journal Science Advances, the paper explains how the dynamics of a polymer called chromatin—the structure into which DNA is packaged—regulate gene expression.
Through the use of machine learning and statistical algorithms, a research team led by Physics Professor Arpita Upadhyaya and National Institutes of Health Senior Investigator Gordon Hager discovered that chromatin can switch between a lower and higher mobility state within seconds. The team found that the extent to which chromatin moves inside cells is an overlooked but important process, with the lower mobility state being linked to gene expression.
Notably, transcription factors (TFs)—proteins that bind specific DNA sequences within the chromatin polymer and turn genes on or off—exhibit the same mobility as that of the piece of chromatin they are bound to. In their study, the researchers analyzed a group of TFs called nuclear receptors, which are targeted by drugs that treat a variety of diseases and conditions.
Recent advancements in deep learning have significantly impacted computational imaging, microscopy, and holography-related fields. These technologies have applications in diverse areas, such as biomedical imaging, sensing, diagnostics, and 3D displays. Deep learning models have demonstrated remarkable flexibility and effectiveness in tasks like image translation, enhancement, super-resolution, denoising, and virtual staining. They have been successfully applied across various imaging modalities, including bright-field and fluorescence microscopy; deep learning’s integration is reshaping our understanding and capabilities in visualizing the intricate world at microscopic scales.
In computational imaging, prevailing techniques predominantly employ supervised learning models, necessitating substantial datasets with annotations or ground-truth experimental images. These models often rely on labeled training data acquired through various methods, such as classical algorithms or registered image pairs from different imaging modalities. However, these approaches have limitations, including the laborious acquisition, alignment, and preprocessing of training images and the potential introduction of inference bias. Despite efforts to address these challenges through unsupervised and self-supervised learning, the dependence on experimental measurements or sample labels persists. While some attempts have used labeled simulated data for training, accurately representing experimental sample distributions remains complex and requires prior knowledge of sample features and imaging setups.
To address these inherent issues, researchers from the UCLA Samueli School of Engineering introduced an innovative approach named GedankenNet, which, on the other hand, presents a revolutionary self-supervised learning framework. This approach eliminates the need for labeled or experimental training data and any resemblance to real-world samples. By training based on physics consistency and artificial random images, GedankenNet overcomes the challenges posed by existing methods. It establishes a new paradigm in hologram reconstruction, offering a promising solution to the limitations of supervised learning approaches commonly utilized in various microscopy, holography, and computational imaging tasks.
A mysterious quantum phenomenon reveals an image of an atom like never before. You can even see the difference between protons and neutrons.
The Relativistic Heavy Ion Accelerator (RHIC), from the Brookhaven Laboratory in the United States, is a sophisticated device capable of accelerating gold ions to a speed of up to 99.995% that of light. Thanks to him, it has recently been possible to verify, for example, Einstein’s famous equation E=mc2.
UC San Diego’s Q-MEEN-C is developing brain-like computers through mimicking neurons and synapses in quantum materials. Recent discoveries in non-local interactions represent a critical step towards more efficient AI hardware that could revolutionize artificial intelligence technology.
We often believe that computers are more efficient than humans. After all, computers can solve complex math equations in an instant and recall names that we might forget. However, human brains can process intricate layers of information rapidly, accurately, and with almost no energy input. Recognizing a face after seeing it only once or distinguishing a mountain from an ocean are examples of such tasks. These seemingly simple human functions require considerable processing and energy from computers, and even then, the results may vary in accuracy.
How close the measured value conforms to the correct value.
A groundbreaking theoretical proof reveals that using a technique called overparametrization enhances performance in quantum machine learning.
Machine learning is a subset of artificial intelligence (AI) that deals with the development of algorithms and statistical models that enable computers to learn from data and make predictions or decisions without being explicitly programmed to do so. Machine learning is used to identify patterns in data, classify data into different categories, or make predictions about future events. It can be categorized into three main types of learning: supervised, unsupervised and reinforcement learning.
In this video I discuss New Cerebras Supercomputer with Cerebras’s CEO Andrew Feldman.
Timestamps:
00:00 — Introduction.
02:15 — Why such a HUGE Chip?
02:37 — New AI Supercomputer Explained.
04:06 — Main Architectural Advantage.
05:47 — Software Stack NVIDIA CUDA vs Cerebras.
06:55 — Costs.
07:51 — Key Applications & Customers.
09:48 — Next Generation — WSE3
10:27 — NVIDIA vs Cerebras Comparison.
Mentioned Papers:
Massively scalable stencil algorithm: https://arxiv.org/abs/2204.03775
https://www.cerebras.net/blog/harnessing-the-power-of-sparsi…-ai-models.
https://www.cerebras.net/press-release/cerebras-wafer-scale-…ge-models/
Programming at Scale:
https://8968533.fs1.hubspotusercontent-na1.net/hubfs/8968533…tScale.pdf.
Massively Distributed Finite-Volume Flux Computation: https://arxiv.org/abs/2304.
Continue reading “This New AI Supercomputer Outperforms NVIDIA! (with CEO Andrew Feldman)” »
A potentially game-changing theoretical approach to quantum computing hardware avoids much of the problematic complexity found in current quantum computers. The strategy implements an algorithm in natural quantum interactions to process a variety of real-world problems faster than classical computers or conventional gate-based quantum computers can.
“Our finding eliminates many challenging requirements for quantum hardware,” said Nikolai Sinitsyn, a theoretical physicist at Los Alamos National Laboratory. He is co-author of a paper on the approach in the journal Physical Review A. “Natural systems, such as the electronic spins of defects in diamond, have precisely the type of interactions needed for our computation process.”
Sinitsyn said the team hopes to collaborate with experimental physicists also at Los Alamos to demonstrate their approach using ultracold atoms. Modern technologies in ultracold atoms are sufficiently advanced to demonstrate such computations with about 40 to 60 qubits, he said, which is enough to solve many problems not currently accessible by classical, or binary, computation. A qubit is the basic unit of quantum information, analogous to a bit in familiar classical computing.
