Cloudian, the innovation leader in enterprise object storage systems, has announced the integration of a plug-in within NVIDIA’s Deep Learning GPU Training System (DIGITS) version 6.1 that allows the DIGITS platform to directly access an S3-compatible storage endpoint. DIGITS users can now easily integrate Cloudian’s S3-compatible HyperStore platform for use with deep learning training libraries.
With this integration, data scientists and researchers using DIGITS can now access Cloudian’s object storage platform as a repository for the vast amounts of data needed to give their deep learning applications the information needed to identify patterns and “learn” the best responses to any given situation.
“This integration will enable AI developers more flexibility in how their applications access and store data, and will lead to faster and more accurate decisions and suggestions,” said Sanjay Jagad, senior director of products and solutions at Cloudian. “The HyperStore platform is the ideal repository for this data, being limitlessly scalable and easy to grow. It’s an ideal technology to help the performance of AI/ML applications, which become more powerful as the amount of data available to them increases.”
NVIDIA DIGITS puts the power of†deep learning†– the use of multiple deep neural networks (DNNs) to help machines make sense of complex data — into the hands of engineers and data scientists. DIGITS can be used to rapidly train DNNs for image classification, segmentation and object detection tasks.
