NORDICS20 - Assets

Deep Learning on AWS

Amazon Web Services Resources EMEA

Issue link:

Contents of this Issue


Page 15 of 50

Amazon Web Services Deep Learning on AWS Page 11 Scalability Both deep learning training and inference workloads have spiky characteristics. There may be a specific time period during the project where you may have to scale specific experiments to hundreds and thousands of instances. This is true for inference and when seasonal spike in inference is expected during special events during the year. Planning and forecasting the high-performance compute required to support experiments and seasonal bursts for training and inference is difficult to plan. AWS services optimized for deep learning solves above challenges faced by deep learning engineers and deep learning scientists. Let us take a closer look at the individual buildings blocks.

Articles in this issue

view archives of NORDICS20 - Assets - Deep Learning on AWS