Lookup tables (LUTs) are frequently used to efficiently store arrays of precomputed values for complex mathematical computations. When used in the context of neural networks, these functions exhibit a lack of recognizable patterns which presents an unusual challenge for conventional logic synthesis techniques. Several approaches are known to break down a single large lookup table into multiple smaller ones that can be recombined. Traditional methods, such as plain tabulation, piecewise linear approximation, and multipartite table methods, often yield inefficient hardware solutions when applied to LUT-based NNs. This paper introduces ReducedLUT, a novel method to reduce the footprint of the LUTs by injecting don't cares into the compression process. This additional freedom introduces more self-similarities which can be exploited using known decomposition techniques. We then demonstrate a particular application to machine learning; by replacing unobserved patterns within the training data of neural network models with don't cares, we enable greater compression with minimal model accuracy degradation. In practice, we achieve up to $1.63\times$ reduction in Physical LUT utilization, with a test accuracy drop of no more than $0.01$ accuracy points.

Graph Convolutional Networks (GCNs) are state-of-the-art deep learning models for representation learning on graphs. However, the efficient training of GCNs is hampered by constraints in memory capacity and bandwidth, compounded by the irregular data flow that results in communication bottlenecks. To address these challenges, we propose a message-passing architecture that leverages NUMA-based memory access properties and employs a parallel multicast routing algorithm based on a 4-D hypercube network within the accelerator for efficient message passing in graphs. Additionally, we have re-engineered the backpropagation algorithm specific to GCNs within our proposed accelerator. This redesign strategically mitigates the memory demands prevalent during the training phase and diminishes the computational overhead associated with the transposition of extensive matrices. Compared to the state-of-the-art HP-GNN architecture we achieved a performance improvement of $1.03\times \sim 1.81\times$.

Apple has released Monterey, the latest version of MacOS. The new software comes with a range of features, many of which have already been seen in iOS 15, which was released last month. They include updates to apps such as Messages and Safari, "Focus" tools that let you block out specific notifications and new FaceTime features including the option to watch films together. That feature, named SharedPlay, has been repeatedly delayed, apparently amid problems having it ready on time. But it now seems to be ready, with Apple announcing that its Fitness service will be taking advantage of it from today.

This week, Apple revealed macOS Monterey, an update to its desktop operating system that will roll out this fall. It comes with a whole host of new features that the company highlighted in its keynote presentation at its annual WWDC event. But if you're stuck on a Mac with an Intel chip, some of those improvements won't make it to your computer at all. Apple computers have been largely powered by Intel chips since 2006. Last November, the company announced that it would switch to its own ARM-based M1 chips. Now with macOS Monterey, Apple has started to leave Intel behind.

Jaguar Land Rover, taking a page from the European luxury car playbook, is offering increasingly attractive performance versions of its entry-level sports cars. One of those was on display at Mazda Raceway Laguna Seca during the recent Monterey Car Week, which culminates at Pebble Beach with the famed Concours D'Elegance. At the track, days before the Concours, Jaguar designers showed off the XE SV Project 8, a street-legal track car, in front of the raceway pits. They boasted deservedly about its good looks and great specifications -- including its 592 horsepower and $187,500 sticker price. You don't have to buy one.

In drone footage captured on May 18 in Monterey, California, a group of orcas is seen carrying out a coordinated attack on a blue whale. Orcas, also known as killer whales, are known to prey on other marine mammals, including dolphins and seals. But even these fearsome predators don't stand much of a chance against a mature blue whale: The largest animal on the planet, an adult blue whale can reach up to a hundred feet long and weigh close to 200 tons. In this instance, the large blue whale flipped on its side, sending up what seemed like a wall of water, and swam away at a speed that far outpaced the orcas, says marine biologist Nancy Black, who captured the event from on board a whale-watching boat. The real reason the orcas likely orchestrated an attack? "They were probably doing it for the heck of it," says Black. "They play with [whales] like cats play with their prey.