"Nvidia shares surge past $700 driven by demand for AI."

LLMs are deep learning layered neural networks of matrices, during forward propagation, inputs are multiplied by weight matrices and the results are added together to produce an output which becomes the next layer's input hence CUDA and floating point arithmetic.

But what if you didn't need multiplication? Seems this news has not reached the market yet. arxiv.org/pdf/2310.11453.pdf

#NVDA #SELL #AI