Achieve up to 6.5x the BERT Deep Learning Performance with AWS M6i Instances Enabled by 3rd Gen Intel® Xeon®Scalable Processors

Description

Bidirectional Encoder Representations from Transformers (BERT) is a framework for natural language processing (NLP) that enables companies to use deep learning for text analysis. Thanks to NLP, these organizations’ applications can predict sentences, suggest answers, and generate chat responses. In third-party testing, AWS EC2 M6i instances enabled by 3rd Gen Intel Xeon Scalable processors delivered significantly better BERT performance than M5a instances enabled by AMD EPYC processors.