Who is using Cerebras chips for training or inference of foundation models at large scale? This seems a bit like a placed puff piece. My understanding is that

Who is using Cerebras chips for training or inference of foundation models at large scale? This seems a bit like a placed puff piece. My understanding is that most of the training and inference happens on Nvidia chips