The representation of nonlinear sub-grid processes, especially clouds, has been a major source of uncertainty in climate models for decades. Cloud-resolving models better represent many of these processes and can now be run globally but only for short-term simulations of at most a few years because of computational limitations. Here we demonstrate that deep learning can be used to capture many advantages of cloud-resolving modeling at a fraction of the computational cost. We train a deep neural network to represent all atmospheric sub-grid processes in a climate model by learning from a multi-scale model in which convection is treated explicitly. The trained neural network then replaces the traditional sub-grid parameterizations in a global general circulation model in which it freely interacts with the resolved dynamics and the surface-flux scheme. The prognostic multi-year simulations are stable and closely reproduce not only the mean climate of the cloud-resolving simulation but also key aspects of variability, including precipitation extremes and the equatorial wave spectrum. Furthermore, the neural network approximately conserves energy despite not being explicitly instructed to. Finally, we show that the neural network parameterization generalizes to new surface forcing patterns but struggles to cope with temperatures far outside its training manifold. Our results show the feasibility of using deep learning for climate model parameterization. In a broader context, we anticipate that data-driven Earth System Model development could play a key role in reducing climate prediction uncertainty in the coming decade.

The implementation of optimal power flow (OPF) methods to perform voltage and power flow regulation in electric networks is generally believed to require communication. We consider distribution systems with multiple controllable Distributed Energy Resources (DERs) and present a data-driven approach to learn control policies for each DER to reconstruct and mimic the solution to a centralized OPF problem from solely locally available information. Collectively, all local controllers closely match the centralized OPF solution, providing near-optimal performance and satisfaction of system constraints. A rate distortion framework facilitates the analysis of how well the resulting fully decentralized control policies are able to reconstruct the OPF solution. Our methodology provides a natural extension to decide what buses a DER should communicate with to improve the reconstruction of its individual policy. The method is applied on both single- and three-phase test feeder networks using data from real loads and distributed generators. It provides a framework for Distribution System Operators to efficiently plan and operate the contributions of DERs to active distribution networks.

The environment is a hot topic, literally. As global temperatures have warmed since 1850, the discussion on what to do about it has heated up as well. Humanity is having an undeniable impact on the natural world. Our growing demand for resources is leading to land-use changes, loss of biodiversity and pollution. Climate change continues to disrupt weather patterns, temperatures and water availability, leading to impacts on human and natural ecosystems -- even the forests are on the move.

AI guides the computers that drive the company's own data centers, helping decide which virtual server to use at any given point in order to complete a task, the company says. That allows it to process cloud tasks more efficiently than a traditional on-premises data center, simply because it's only using the power it needs to get a job done, says Liz Willmott, carbon program lead at Microsoft. Aside from the sheer energy demands of server and computing hardware, cloud storage centers come with hefty cooling costs. Microsoft says that its cloud facilities have been 100% carbon neutral since 2012, and the company has committed to using more renewable energy sources like wind, hydropower, and solar. But it's also been touting the benefits of software and AI for further cutting electricity use.

Later this month, when a ranking of supercomputers called the TOP500 comes out, the highest spot is expected to go to an American machine called Summit. If you wanted to put Summit on your desk, you'd need a workstation that's about the size of two tennis courts. That's because this particular machine occupies 5,600 square feet of space. Its guts spread across more than 250 cabinets, each about the size of a refrigerator. And within the system is 185 miles of fiber optic cables, about enough to stretch from New York City to Providence, Rhode Island.

IBM, NVIDIA, and the U.S. Department of Energy (DOE) recently announced that they have completed testing the world's fastest supercomputer, Summit, at the Oak Ridge National Laboratory in Oak Ridge, Tennessee. Capable of over 200 petaflops (200 quadrillion operations per second), Summit consists of 4600 IBM dual socket Power 9 nodes, connected by over 185 miles of fiber optic cabling. Each node is equipped with 6 NVIDIA Volta TensorCore GPUs, delivering total throughput that is 8 times faster than its predecessor, Titan, for double precision tasks, and 100 times faster for reduced precision tasks common in deep learning and AI. China has held the top spot in the Top 500 for the last 5 years, so this brings the virtual HPC crown home to the USA. Figure 1: The Summit Supercomputer at the Department of Energy's Oak Ridge National Labs is now the fastest computer in the world. Some of the specifications are truly amazing; the system exchanges water at the rate of 9 Olympic pools per day for cooling, and as an AI supercomputer, Summit has already achieved (limited) "exascale" status, delivering 3 exaflops of AI precision performance.

IBM and the Department of Energy's Oak Ridge National Laboratory have revealed the world's "most powerful and smartest scientific supercomputer." Known as Summit, IBM says that its new computer will be capable of processing 200,000 quadrillion calculations per second. To put that into perspective, if every person on Earth did a single calculation per second, it would take 305 days to do what Summit does in a single second. Assuming those numbers are accurate, that would make Summit the world's fastest supercomputer. It would also mark the first time since 2012 that a U.S. computer held that title.

You say you're a parent or teacher investigating robot kits for children? And you don't want a simple solution with a single purpose: you want the child to experience science, technology, engineering, and math? You want a kit that teaches all four categories, from piecing together the foundation to wiring the appendages to programming the "brain" using software. That's where our list of robot kits for kids comes in. Most of the robot kits listed below are tied to terms such as STEM, Arduino, and Blockly.

In sub-Saharan Africa, home electricity is a 50-50 prospect and bank accounts can be rare, but most people have some kind of cellphone. The phones provide information often tough to come by in rural areas--the latest commodity prices, for example. And even in places where pastoral tribesmen tend livestock in very old-school ways, they may also chat over WhatsApp and use money-transfer apps to settle debts. To charge the phones without access to an electrical grid, Africans spend more than $17 billion a year on such fuels as kerosene and firewood to power sometimes primitive generators. Simon Bransfield-Garth is pitching a cleaner and, he says, smarter alternative.

This dissertation investigates the use of one-sided classification algorithms in the application of separating hazardous chlorinated solvents from other materials, based on their Raman spectra. The experimentation is carried out using a new one-sided classification toolkit that was designed and developed from the ground up. In the one-sided classification paradigm, the objective is to separate elements of the target class from all outliers. These one-sided classifiers are generally chosen, in practice, when there is a deficiency of some sort in the training examples. Sometimes outlier examples can be rare, expensive to label, or even entirely absent. However, this author would like to note that they can be equally applicable when outlier examples are plentiful but nonetheless not statistically representative of the complete outlier concept. It is this scenario that is explicitly dealt with in this research work. In these circumstances, one-sided classifiers have been found to be more robust that conventional multi-class classifiers. The term "unexpected" outliers is introduced to represent outlier examples, encountered in the test set, that have been taken from a different distribution to the training set examples. These are examples that are a result of an inadequate representation of all possible outliers in the training set. It can often be impossible to fully characterise outlier examples given the fact that they can represent the immeasurable quantity of "everything else" that is not a target. The findings from this research have shown the potential drawbacks of using conventional multi-class classification algorithms when the test data come from a completely different distribution to that of the training samples.