By Amanda Holpuch , The New York Times.  Excerpt: Manufacturers will no longer get a credit toward vehicle emissions standards by installing engines that automatically stop at red lights. ...The start-stop feature is meant to save fuel and reduce emissions, but the Trump administration rejected the scientific finding that the government used to support vehicle emission reduction regulations, making it possible to eliminate the credit. ...Research shows that start-stop reduces fuel use and cuts emissions. Depending on driving conditions, stop-start improved fuel economy between 7.27 and 26.4 percent during testing, according to a  2023 technical paper  by SAE International, an organization formerly known as the Society of Automotive Engineers....  Full article at https://www.nytimes.com/2026/02/15/business/energy-environment/epa-tax-credits-stop-start-ignition-cars.html . 

By Hannah Richter , Science.  Excerpt: Every time people ask ChatGPT for help, their request percolates through a whirring farm of computers kept cool inside a windowless warehouse. These facilities, known as data centers, gobbled up  more than 4% of U.S. electricity in 2023 . ...researchers reported online earlier this month in  Solar Energy  that they had come up with a clever efficiency boost. By using the Sun’s warmth to raise the temperature of vented waste heat, data center operators can  generate a significant fraction of the electricity  they need while recycling some power....  Full article at https://www.science.org/content/article/solar-bump-could-help-data-centers-recover-wasted-energy . 

By Stephen Proctor, The Cool Down (TCD).  Excerpt: Two innovative companies, Solarwatt and Stiebel Eltron U.K., have collaborated on a first-of-its-kind home in the United Kingdom. According to  Solarwatt , the companies unveiled their fully integrated solar energy and  heat pump  system at a home in North Wales....  Full article at https://www.thecooldown.com/green-home/solarwatt-stiebel-eltron-uk-home-energy/ . 

By Emily Cerf , UC Santa Cruz News Center.  Excerpt: Large language models such as ChaptGPT have proven to be able to produce remarkably intelligent results, but the energy and monetary costs associated with running these massive algorithms is sky high. It costs $700,000 per day in energy costs to run ChatGPT 3.5, according to recent estimates, and leaves behind a massive carbon footprint in the process. In a  new preprint paper , researchers from UC Santa Cruz show that it is possible to eliminate the most computationally expensive element of running large language models, called matrix multiplication, while maintaining performance. In getting rid of matrix multiplication and running their algorithm on custom hardware, the researchers found that they could power a billion-parameter-scale language model on just 13 watts, about equal to the energy of powering a lightbulb and more than 50 times more efficient than typical hardware....  Full article at https://news.ucsc.edu/...