Hewlett Packard Enterprise on Wednesday showcased machine learning platforms it says can enable users to develop and train AI models faster and at greater scale.
HPE’s machine learning development system integrates a machine learning software platform, GPU and CPU chips and accelerators, and networking to enable users to reduce the time they spend creating and train machine learning models from weeks and months to days, according to HPE.
HPE Swarm Learning, on the other hand, is a machine learning framework for edge or distributed sites.
Since most of the training of AI models happens in a central location, developers and IT professionals have to move large volumes of data between repositories and this constant exchange of data is further slowed down by privacy and data ownership requirements, according to HPE.
Swarm Learning allows users to train and operate models at the edge rather than at the server level. The HPE AI system also enables users to work with distributed data and build machine learning models, while maintaining data governance and privacy, the vendor said.
The old and the new
Machine Learning Development System and Swarm Learning have been in the works for some time, said Evaluator Group analyst Frederic Van Haren.
The vendor has started discussing its plans for HPE Swarm, especially in 2020, he said. Now, the product’s focus on edge and blockchain security could help it gain traction with organizations looking for these components.
HPE’s machine learning development system stems from its acquisition of Determined AI in 2021, but “I don’t see the value HPE adds,” Van Haren said.
Determined AI provided an open-source deep learning platform on which data scientists can train models and share GPU resources.
The machine learning development system enables HPE to combine determined AI software with HPE hardware, the vendor said.
The system includes a comprehensive software stack including purpose-built artificial intelligence tools now called the HPE Machine Learning Development Environment, as well as the vendor’s Docker container technology, HPE Performance Cluster Manager, and the Red Hat Enterprise Linux operating system.
While both offerings could attract the attention of government entities, labs and large organizations, Van Haren said he doesn’t see the vendor garnering much interest from businesses.
“HPE is known for selling hardware, not software or solution stacks,” he said.
Late to market
HPE’s products may not be competitive in the machine learning market, said Andy Thurai, analyst at Constellation Research.
“HPE is late to the market with many players in this market already,” he said.
Although the HPE Apollo 6500 Gen10 System is powerful enough to support faster creation and training of machine learning models, and distributed training and automated hyperparameter tuning allow developers to train models more efficiently, the machine learning development system “can only compete [clouds] which is somewhat limited target space,” Thurai said.
“Most companies have already adopted one of the public clouds for their model training,” he said, referring to AWS, Azure or Google. “It will be an uphill battle trying to bring him back in-house.”
However, Swarm Learning may have some potential with businesses, Thurai said.
“While current ML model training and consumption needs to change to use this technology, there is potential for it,” he said.
The AI tool allows businesses to skip the lengthy process of redacting, cleaning, protecting and managing data before moving it to a particular location.
“The distributed ML model framework can be particularly useful in situations where edge locations may have volumes of datasets. Otherwise, it will be useless,” he said.
According to HPE, the machine learning development system is priced on a cluster basis. The provider has not yet made pricing information available for Swarm Learning. Both AI systems are available now.
Comments are closed.