Using Bidirectional Encoder Representations from Transformers (BERT) machine learning helps organizations make sense of large amounts of text. Speeding up this analysis helps teams gain insights from that data faster. Principled Technologies (PT) tested the BERT performance of three AWS instance types: M6i instances with 3rd Gen Intel Xeon Scalable processors, M5n instances with 2nd Gen Intel Xeon Scalable processors, and M6a instances with 3rd Gen AMD EPYC processors.
According to the report, “In tests at multiple instance sizes, AWS M6i instances offered up to 45 percent better BERT performance on a benchmark from the Intel Model Zoo than the M5n instances with previous-gen processors and up to 6.4 times the BERT performance compared to M6a instances with 3rd Gen AMD EPYC processors.”
To learn more about how AWS M6i instances performed in BERT testing, read the report at https://facts.pt/HnKi6oO.
For more information, please visit www.principledtechnologies.com.