Lifeboat Foundation

Nvidia today announced the release of TensorRT 8, the latest version of its software development kit (SDK) designed for AI and machine learning inference. Built for deploying AI models that can power search engines, ad recommendations, chatbots, and more, Nvidia claims that TensorRT 8 cuts inference time in half for language queries compared with the previous release of TensorRT.

Models are growing increasingly complex, and demand is on the rise for real-time deep learning applications. According to a recent O’Reilly survey, 86.7% of organizations are now considering, evaluating, or putting into production AI products. And Deloitte reports that 53% of enterprises adopting AI spent more than $20 million in 2019 and 2020 on technology and talent.

TensorRT essentially dials a model’s mathematical coordinates to a balance of the smallest model size with the highest accuracy for the system it’ll run on. Nvidia claims that TensorRT-based apps perform up to 40 times faster than CPU-only platforms during inference, and that TensorRT 8-specific optimizations allow BERT-Large — one of the most popular Transformer-based models — to run in 1.2 milliseconds.

A new set of equations can precisely describe the reflections of the Universe that appear in the warped light around a black hole.

The proximity of each reflection is dependent on the angle of observation with respect to the black hole, and the rate of the black hole’s spin, according to a mathematical solution worked out by physics student Albert Sneppen of the Niels Bohr Institute in Denmark.

This is really cool, absolutely, but it’s not just really cool. It also potentially gives us a new tool for probing the gravitational environment around these extreme objects.

Circa 2019

MIT’S new mini cheetah robot is the first four-legged robot to do a backflip. At only 20 pounds the limber quadruped can bend and swing its legs wide, enabling it to walk either right side up or upside down. The robot can also trot over uneven terrain about twice as fast as an average person’s walking speed. (Learn more: http://news.mit.edu/2019/mit-mini-cheetah-first-four-legged-…kflip-0304)

Continue reading “Backflipping MIT Mini Cheetah” »

A team of researchers affiliated with multiple institutions in China, working at the University of Science and Technology of China, has achieved another milestone in the development of a usable quantum computer. The group has written a paper describing its latest efforts and have uploaded it to the arXiv preprint server.

Back in 2019, a team at Google announced that they had achieved “quantum supremacy” with their Sycamore machine—a 54 qubit processor that carried out a calculation that would have taken a traditional computer approximately 10000 years to complete. But that achievement was soon surpassed by other teams from Honeywell and a team in China. The team in China used a different technique, one that involved the use of photonic qubits—but it was also a one-trick pony. In this new effort, the new team in China, which has been led by Jian-Wei Pan, who also led the prior team at the University of Science and Technology has achieved another milestone.

The new effort was conducted with a 2D programable computer called Zuchongzhi—one equipped to run with 66 qubits. In their demonstration, the researchers used only 56 of those qubits to tackle a well-known computer problem—sampling the output distribution of random quantum circuits. The task requires a variety of computer abilities that involve mathematical analysis, matrix theory, the complexity of certain computations and probability theory—a task approximately 100 times more challenging than the one carried out by Sycamore just two years ago. Prior research has suggested the task set before the Chinese machine would take a conventional computer approximately eight years to complete. Zuchongzhi completed the task in less than an hour and a half. The achievement by the team showed that the Zuchongzhi machine is capable of tackling more than just one kind of task.

Quantum computing algorithms can simulate infinitely-large quantum systems thanks to mathematical tools known as tensor networks.

Math about black holes:

If you’ve been following the arXiv, or keeping abreast of developments in high-energy theory more broadly, you may have noticed that the longstanding black hole information paradox seems to have entered a new phase, instigated by a pair of papers [1, 2] that appeared simultaneously in the summer of 2019. Over 200 subsequent papers have since appeared on the subject of “islands”—subleading saddles in the gravitational path integral that enable one to compute the Page curve, the signature of unitary black hole evaporation. Due to my skepticism towards certain aspects of these constructions (which I’ll come to below), my brain has largely rebelled against boarding this particular hype train. However, I was recently asked to explain them at the HET group seminar here at Nordita, which provided the opportunity (read: forced me) to prepare a general overview of what it’s all about. Given the wide interest and positive response to the talk, I’ve converted it into the present post to make it publicly available.

Continue reading “Islands behind the horizon” »

A group of mathematicians has shown that at critical moments, a symmetry called rotational invariance is a universal property across many physical systems.

Circa 2014

Physicists have verified a key prediction of Albert Einstein’s special theory of relativity with unprecedented accuracy. Experiments at a particle accelerator in Germany confirm that time moves slower for a moving clock than for a stationary one.

The work is the most stringent test yet of this ‘time-dilation’ effect, which Einstein predicted. One of the consequences of this effect is that a person travelling in a high-speed rocket would age more slowly than people back on Earth.

Continue reading “Einstein’s ‘Time Dilation’ Prediction Verified” »

Nathan Seiberg, 64, still does a lot of the electrical work and even some of the plumbing around his house in Princeton, New Jersey. It’s an interest he developed as a kid growing up in Israel, where he tinkered with his car and built a radio.

“I was always fascinated by solving problems and understanding how things work,” he said.

Seiberg’s professional career has been about problem solving, too, though nothing as straightforward as fixing radios. He’s a physicist at the Institute for Advanced Study, and over the course of a long and decorated career he has made many contributions to the development of quantum field theory, or QFT.