@aidailypost.comEvery year MIT Lincoln Lab spots 5‑10 fresh AI chip startups—think new ASICs, GPUs and neural‑net accelerators. The race to power next‑gen ML is heating up. Curious? Dive into the full breakdown. #AIchips #AIaccelerators #MachineLearning
@kit.edu42,000 #ProcessorCores, 300 #AIAccelerators, and a planned peak performance of around 33 #petaFLOPS: #KITKarlsruhe puts new #supercomputer into operation. #HoreKa 2 will support researchers in their efforts to address complex scientific issues more quickly.
www.kit.edu/kit/english/...
@michel-k-tech.bsky.socialSamsung's zHBM roadmap pushes HBM from a bandwidth booster toward tighter compute integration, which could matter a lot for future AI accelerators.